AMD's 64-bit Plot

ceebABC writes "In a long interview with eWEEK, AMD's CEO Hector de Ruiz talks about struggling to compete with Intel, but more importantly about their upcoming 64-bit processors. He says that AMD's 64-bit chips will be comparatively priced to the 32-bit ones, and backwards compatible. He also thinks there will be a market for desktop 64-bit systems. Skip to the last page for the most interesting stuff."

Comments

Big deal.

[Anonymous Coward]

Nintendo64 was 64 bits.

Re:Big deal.

[Devil's BSD]

Nintendo64 was 64 bits.

Yeah, and I have a 128-bit graphics card. (I know, they have like 100 Mbit ethernet cards now. :) ) However, The GPU and processor are totally different. The graphics card has more bits but obviously it doesnt run as fast as the cpu. All it does is make your fragfest a little more purty by letting you see the giblets all over. Having the CPU 64 bits is quite different, security-wise, code-wise, and speed-wise. If you have a 64-bit 2 GHz processor and a 32-bit 2 GHz processor, the 64-bit processor is going to be much faster. This speeds up the whole system, not just the rate at which you make giblets fly.

Re:Big deal.

[Anonymous Coward]

I know, they have like 100 Mbit ethernet cards now.

Umm, well there is the 1Gbit NICs.

I know, they're a bit expensive, but they do exist.

Re:Big deal.

[Isle]

If you have a 64-bit 2 GHz processor and a 32-bit 2 GHz processor, the 64-bit processor is going to be much faster. This speeds up the whole system, not just the rate at which you make giblets fly.

Ehrmm. no, if it were that easy we would all be using 64bit by now. 64bit has historically been faster because they belong to a better group of architectures called RISC, the new AMD 64-bit will be faster not because they have more bits but because AMD has upgraded the architecture and added more registers.

The number of bits is a meaningless as counting the number of seats in a car, twice as many seats doesnt make a faster car. In fact it makes the car harder to design to be fast, so does 64bit processors.

Re:Big deal.

[Junks Jerzey]

If you have a 64-bit 2 GHz processor and a 32-bit 2 GHz processor, the 64-bit processor is going to be much faster. This speeds up the whole system, not just the rate at which you make giblets fly.

No. That's a myth. As it stands, Pentiums for many years now have sported 64 bit buses and 64-bit FPUs (well, 80-bit CPUS actually), so we're not talking about bus size and FPU width. We're talking about:

1. All addresses being 64-bits.
2. All internal integer registers being 64-bits.

For #1, realize that this is going to greatly increase the data size of many applications. The larger the data size, the higher the chance of cache misses. In general, this is a loss, not a win.

For #2, realize that some integer operations are O(N) where N is the number of bits involved. 64-bit multiplication and division are slower than the same 32-bit operations. Period.

The gain with 64-bit processors is one of address space and nothing more.

Re:Big deal.

[Anonymous Coward]

RE: "If you have a 64-bit 2 GHz processor and a 32-bit 2 GHz processor, the 64-bit processor is going to be much faster."

That is not true. What is your argument for this?
It sounds like you mean in an otherwise equal machine, for a start you are going to have cache problems. Many 64-bit machines use 32-bit instruction formats, but data will be 64-bits and obviously fewer words fit in the data cache than with 32-bit data words. Potentially, that alone can give you a massive slowdown.

Please supply your reasoning for this claim..

Re:Big deal.

[Anonymous Coward]

OK, a comparison of 32 and 64 bit processors running at the same speed would probably NOT show that much of a difference...
Assuming that the memory subsystem is the same then the 64-bits processor would only have an edge when:
a) Using lots of address space.
b) Doing 64 bits integer arithmetic (floting point arith is the same for 32 and 64).
... there might be other areas where 64 bits are better. But 'speed-wise' is not one of them.
'Code-wise' probably isn't either as most 64 bit'ers use 32 bit instruction words.

I actually read a primer on AA-64 and instruction words are still 32 bits (actually they are variable length) and the fpu doesn't seem to benefit from 64 bits in the integer units.

/AC doing 64b CPU design for food.

Note: What is this stuff about 128 bit graphics? N64 actually had a 64 bit CPU (MIPS).

Re:Big deal.

[Anonymous Coward]

So was the Atari Jaguar, but nobody knows about that system. It came out years before Nintendo made the Nintendo Ultra 64.

Re:Big deal.

[rock_climbing_guy]

no, the Commodore 64 was 64 bi^h^hkilobytes of RAM. hey, imagine a beowulf cluster of those.

Re:Big deal.

[Dave_bsr]

Increased maximum memory helps.
Opteron's extra registers help.
64-bit calculations are easier, they don't have to be put into multiple 32-bit parts.

So...a 32-person bus is just as good as a 64-person bus? It may be harder to design and build, but when you have to move >32 people it's nice to have that big of a bus running around.

What I'm saying is, being 64-bit DOES make you faster. Not twice as fast, but definately faster and more powerful.

Re:Big deal.

[Mitreya]

For #1, realize that this is going to greatly increase the data size of many applications. The larger the data size, the higher the chance of cache misses. In general, this is a loss, not a win.

wouldn't the chance of cache misses depend on the caching policy? How does the data size matter? If your policy is good, when your misses will be rare. Otherwise you're screwed even if it is 8-bit

The gain with 64-bit processors is one of address space and nothing more.

Which includes better behaviour for those programs that have to fake larger address space. That would be a speed increase.

Re:Big deal.

[Christopher Thomas]

For #2, realize that some integer operations are O(N) where N is the number of bits involved. 64-bit multiplication and division are slower than the same 32-bit operations. Period.

If area is not an issue, you can do a multiplication in log N steps. Current latencies suggest that this has been done for a while, so I wouldn't expect to see much of a difference at all for multiplying numbers. It's division that takes O(N) steps (you can apply brute force to reduce the proportionality constant, but the order remains).

Re:Big deal.

[cioxx]

Having the CPU 64 bits is quite different, security-wise...

Would you mind elaborating on that part?

Re:Big deal.

[styrotech]

The number of bits is a meaningless as counting the number of seats in a car, twice as many seats doesnt make a faster car. In fact it makes the car harder to design to be fast, so does 64bit processors.

It does let you transport people twice as fast though - although most trips will still end up being made with some empty seats.

I'll leave the car pooling analogy to someone else :)

Re:Big deal.

[Christopher Thomas]

For #1, realize that this is going to greatly increase the data size of many applications. The larger the data size, the higher the chance of cache misses. In general, this is a loss, not a win.

wouldn't the chance of cache misses depend on the caching policy? How does the data size matter?

Data size matters because a program will typically access a fixed number of working variables, not a fixed amount of data. If a program's working set size stays at, say, 1000 words, and you move from a 32-bit to a 64-bit architecture, you need a cache with twice as much storage space to hold the working set without thrashing.

There's easily enough die area to double the sizes of the L1 and L2 caches; the problem is that it slows down cache access (more latency cycles fetching something from L1 is a Bad Thing).

Certain types of load work with constant size instead of constant word count, but most of those deal with working sets large enough that you'll thrash no matter what.

The gain with 64-bit processors is one of address space and nothing more.

Which includes better behaviour for those programs that have to fake larger address space. That would be a speed increase.

Nothing running on x86 will do that. Unless you're running old DOS programs in real mode, you're already working with a flat address space. Typically 2 gigs of this is available to user programs (with the rest being mapped to kernel or device space). If you have a problem with a working set larger than 2 gigabytes, you already have a Sun/$other_vendor machine to solve it on.

Larger address space targets the _future_ problem of desktop users who want many gigabytes of memory.

A fringe benefit is being able to more efficiently map multi-gigabyte files into memory space, but performance for this kind of task is limited by disk latency and controller bandwidth, not memory architecture.

Re:Big deal.

[Anonymous Coward]

WTF is up with eWeek's blue stripe in the middle of the page. I can't read the fucking text!! WTF!!

Re:Big deal.

[halo8]

This is a Question ive wanted answered for a while now..
these news 64bit proc's
will they bis SISC? or RISC?
isnt it time we dumped the whole SISC architecture? like the ISA standard

Re:Big deal.

[Anonymous Coward]

You can use Newton's method. Combined w/ constant time multiplication, division is only O(log N)

Re:Big deal.

[Horn]

Seeing as the hammer is backwards compatable with x86 the cpu is a SISC design.

Re:Big deal.

[joto]

isnt it time we dumped the whole SISC architecture? like the ISA standard

As for as I know, the SISC (single instruction set computing), typically embodied by the instruction SBN (subtract and branch if negative) is only used as a joke, in the same manner as Intercal and Malbolge.

Oh, you probably meant CISC, never mind...

Re:Big deal.

[Gavitron_zero]

The number of bits is a meaningless as counting the number of seats in a car, twice as many seats doesnt make a faster car. In fact it makes the car harder to design to be fast, so does 64bit processors

That's not exactly accurate. A 64 bit processor has a large data pathway, and is more comparable to a roadway than a car. The cars are the data, and a 64-bit roadway has twice the space for cars (data) on it, which is where the extra speed is. But I do agree with you otherwise.

Re:Big deal.

[Christopher Thomas]

You can use Newton's method. Combined w/ constant time multiplication, division is only O(log N)

Newton's Method involves doing a division...

Re:Big deal.

[Anonymous Coward]

Looks fine to me. Try upgrading your browser to one that doesn't suck as much such as Mozilla.

Re:Big deal.

[Directrix1]

CISC

Re:Big deal.

[Anonymous Coward]

But if your an atheist aren't you a heretic since your a non-believer of the Catholic Church, that's what a heretic is a non-believer and an atheist is a non-believer.

Re:Big deal.

[ameoba]

I think you mean CISC.

RISC = Reduced Instruction Set Computer
CISC = Complex ...

The basic idea of (most) RISC chip designs, such as the MIPS, Alpha, PowerPC & Sparc, was to have a large number of general purpose registers, fixed length instructions that could only refer to those registers, and only a handful of instructions that specifically read/wrote to main memory (which is why they're also referred to as 'load/store' architectures). This simplistic design allowed them to push clock speeds without too much trouble. RISC processors were also adopted superscalar designs (having multiple execution units, allowing the execution of multiple instructions 'simultaniously') before their CISC counterparts.

In contrast to the simplicity of the RISC systems, there are the CISC chips, such as the x86 and the old VAX processors, which tried to make their instructions resemble high-level languages, as well as having a smaller number of registers, many of them having a special purpose. With variable length instructions, and many different modes of operation for each instruction, the CISC methodology generaly resulted in much larger, more complex chip designs that were harder to speed up, pipeline & make superscalar.

To compare the two, lets take a simple operation, such as taking two numbers from memory & adding them together. A generic RISC system would do something like:
1) load 1st number into Register 1
2) load 2nd number into Register 2
3) add the value in R1 to R2, putting the value in R3
4) copy the value from Register 3 to memory ...and not have any other way to solve the problem

where a CISC chip, would more likely do something more like:
1)add the value at memory location 1 to the value at memory location 2, and store in a special Accumulator register
2) copy the Accumulator register back to memory

The difference being that where the RISC machine only had one addition operation (register+register->register), the CISC machine would have a handful of them, depending on where the data came from (memory (using multiple forms of reference), registers, constants, and various combinations).

In the early 80s, the RISC/CISC debate was a hot one in accademia, and RISC won out there, by virtue of its simplicity & easy of improvement. By the mid 80s, the debate was starting again in industry, as a number of RISC chips started entering the marketplace, where Intel's x86 architecture won by virtue of the IBM PC.

The whole debate is pretty much a moot point now,
since Intel's new x86 chips have RISC cores wrapped by a thin layer to translate the complex instructions. As an added bonus, the new 64b x86 systems should be adding a bunch of extra registers, further negating the penalty of the architecture.

Re:Big deal.

[Ninja Programmer]

You are thinking about large integers. Actually at the circuit level, multiplication is not done in O(log(N)) steps, but instead, they just have blazingly fast adders. Now, of course the additions can be grouped in a parallel tree, so there are essentially log(N) stages. Each adder can be thought of as O(log(N)), so you might say that the multiply is O((log(N))^2) steps.

Re:Big deal.

[Anonymous Coward]

Another benefit is the increased maximum amount of ram and the increase in speed since the processor will also feature a memory controller onboard.

Re:Big deal.

[Christopher Thomas]

You are thinking about large integers. Actually at the circuit level, multiplication is not done in O(log(N)) steps, but instead, they just have blazingly fast adders.

I have been treating adders as operating in constant time, as we're working with small integers.

The grid approach to multiplication that everyone sees in undergrad microprocessor design requires N serial additions. A tree structure still requires N additions, but done in parallel in log(N) levels (as you pointed out). This is where my statement came from.

Very large integers would indeed require O(log(N)^2) cycles, due to non-negligeable adder delay, but we can make the constant-time approximation for our purposes (especially since the pipeline is likely tuned to have adds take one cycle).

Re:Big deal.

[Anonymous Coward]

But what if you only want to move only less than 32 people, then your hauling extra space around. 64-bit processo is much slower than 32-bit until you start moving more than 2GB for each process. Thats why most 64-bit workstations also do 32-bits too, and also the only reason why almost all programs for these machines are compiled in 32-bit instead of 64 (besides compatibility).

Re:Big deal.

[Weirsbaski]

1. All addresses being 64-bits.
2. All internal integer registers being 64-bits.

For #1, realize that this is going to greatly increase the data size of many applications. The larger the data size, the higher the chance of cache misses. In general, this is a loss, not a win.

For #2, realize that some integer operations are O(N) where N is the number of bits involved. 64-bit multiplication and division are slower than the same 32-bit operations. Period.
The gain with 64-bit processors is one of address space and nothing more.

You do realize that x86 and x86-64 let you choose the data-size on an instruction-by-instruction basis, don't you? So if you want to have a huge table of 32-bit (or 16-bit or 8-bit) values, you can still do that. Or if you only need 32-bit multiplies and divides, you can still do that too.

The 64-bit extensions are just that- extensions. Hammer can still do everything that Athlon/Pentium can do.

Re:Big deal.

[Anonymous Coward]

Exactly, that is why you need a competent scheduler and people to get on and take advantage of our wonderous new transportation system! ;-)

Re:Big deal.

[Bert64]

The atari Jaguar had a 64bit graphics processor, just like a 486 does if you plug an S3 Trio64 card in...
The main processor of the jaguar was a Motorola 68000, a 32bit processor with 16bit external paths and 24bit addressing.

Re:Big deal.

[Anonymous Coward]

Try as I might, I find it hard to find a browser that sucks as much as mozilla.

Opera - load times for impatient people.

Re:Big deal.

[Anonymous Coward]

[inserts curved index finger into mouth]

FISHED IN!! FISHED IN!!

sucker ;-)

Re:Big deal.

[Halo1]

Increased maximum memory helps.

But if you don't need the extra memory, all you get is the overhead of having to load 64bit pointers all the time instead of 32bit ones. This means that you have more memory->processor traffic, slowing down things instead of speeding them up.

Opteron's extra registers help.

This doesn't have anything to do with 64bit.

64-bit calculations are easier, they don't have to be put into multiple 32-bit parts.

Most integer calculations however don't need 64bit precision, 32bits is often enough.

Really, apart from the increased memory address space, there aren't many advantages (which result in speedups) that 64bit processors hold compared to 32bit ones. And even that extra memory space is only useful to a select number of applications (such as 3D renderers and scientific applications) currently, although I suppose that will change over time.

Re:Big deal.

[Anonymous Coward]

>Data size matters because a program will typically >access a fixed number of working variables, not a >fixed amount of data. If a program's working set >size stays at, say, 1000 words, and you move from a >32-bit to a 64-bit architecture, you need a cache >with twice as much storage space to hold the >working set without thrashing.

Infact thrashing occurs when an OS is unable to keep a large enough RMS(Resident Memory Size) of a process. It has *nothing* to do with the cache.
So what happens is that the OS must swap pages from secondary memory(Virtual RAM) into main memory(Physical RAM)- when a lot of this is going on then this is termed _thrashing_.

Re:Big deal.

[Keith_Beef]

even that extra memory space is only useful to a select number of applications

One word: ramdisk.

Here's a first example of what you can do. DVD or CD-Rom is very slow, hard drive space is faster and is cheap. So I copy the contents of my reference books from the DVD onto the hard drive. But this too is slow. So when my machine boots, I would like to be able to load a ramdisk image of two or three DVD's worth of reference material.

A second example, would be to do the same, but with something like the contents of /bin, /sbin, /usr/bin and /usr/sbin to speed up the execution of my bash and ksh scripts.

Having something like six gigabytes of memory would make this perfectly feasible.

Most integer calculations however don't need 64bit precision, 32bits is often enough.

That reminds me of the 640kByte anecdote.

Re:Big deal.

[Halo1]

One word: ramdisk.

[snip]

Having something like six gigabytes of memory would make this perfectly feasible.

I'm not going to nitpick about how you're going to cram the contents of 3 DVD's in 6 GB, but in the days of DOS/Win 3.1 when people also had 8 or 16 MB of ram, they also didn't do this. At most, they used a ram disk for temp space. Nowadays, ram disks are even less usefull since the OS'es have much more intelligent caching functionality (like using all otherwise unused RAM as disk cache). So, yes, you could get a 4GB disk cache now, which will speed up your system.

I don't see this immediately happening in consumer (or even prosumer) systems however, because you'll either need a lot of dimm slots or very high density dimms and both will be quite expensive.

Please also note that I was reacting to the original poster's statement that 64bit processing would automatically make your system faster. That's simply not true. If you start adding lots of ram, then it will become faster because of the ram (which is indeed made possible because you have 64 bit address space, although like other have already said in this thread, the address bus is often already larger than 32 bits in the 32 bit processors we use today).

Most integer calculations however don't need 64bit precision, 32bits is often enough.

That reminds me of the 640kByte anecdote.

This was also in reaction to the statement that 64 bit processing would automatically make your system faster. If lots of programs had to do lots of 64bit calcuations, then this would be true. The fact is that very little calcualtions require 64 bit integer precision. Nothing says this won't change in the future, but until then your programs will not run any faster on a 64 bit processor (even when completely recompiled) than on a 32 bit processor. In fact, they'll most likely be slower because of the extra memory bandwidth that's necessary for loading pointers.

Re:Big deal.

[fredrik70]

I think he used the word thrashing a bit 'loose' there, but you can almost see it as thrashing from point of view of the CPU.

Re:Big deal.

[vile maxim]

using mozilla... having no problem! Try again Mr. Anonymous. -Jeff

Re:Big deal.

[jpc]

> For #2, realize that some integer operations are
> O(N) where N is the number of bits involved.
> 64-bit multiplication and division are slower
> than the same 32-bit operations. Period.

Not true. They are done in circiuts so you can use parallel algorithms. Mostly arithmetic ops are n log n so they dont take a significant amount of time longer. You do need more transistors though...

Re:Big deal.

[Christopher Thomas]

Infact thrashing occurs when an OS is unable to keep a large enough RMS(Resident Memory Size) of a process. It has *nothing* to do with the cache.

In Computer Engineering, "thrashing" refers to the overflow of any kind of cache, be it a CPU's cache of system RAM or system RAM's cache of the OS's virtual memory space (RAM plus swap plus what-have-you).

64 bits

[scrawny]

that's only $8. how can they be as cheap as a 32 bit computer?

Re:64 bits

[Anonymous Coward]

Oh please.. it might have not been the best joke, but there was no need to mod it down.

Hmm

[zapfie]

Is there really much consumer (not business) application for 64-bit processors? If so, where would desktop computing benefit?

Re:Hmm

[Anonymous Coward]

Games of course. First come games, then comes your photo scanning software, then comes your word processor.

Re:Hmm

[nmaeone]

Well, 64 bits made Mario come to life. Maybe Micro$oft will make a fully 3-D rendered MSN Butterfly to help you with your daily tasks?

Re:Hmm

[Anonymous Coward]

Clip (or Clippy?) will do fine, thank you very much.

Re:Hmm

[MisterFancypants]

There's really not much need for 64bits even in gaming. The side-effects of going to these 64 bit chips (Hammer and Itanium), like executing more instructions per cycle, will help out, but the 64-bitness of the chips is not at all important for games for the foreseeable future.

Re:Hmm

[rodgerd]

Desktop video is one that springs to mind - lotsa memory.

I can buy PC133 @ US$60 per half gig. For US$500, I can fill the address space of the 32 bit processor, yet a non-trivial home movie could occupy more than 4 GB in uncompressed form.

Re:Hmm

[Anonymous Coward]

Hmm, being able to execute more instructions per cycle isn't really going to help you if there is not significant instruction level parallelism to be exploited! Limited reoder window size on the processor is a big limit here, as is the complexity of control logic for wide OOO. One big benefit of x86-64 is simply that in long mode the ISA is much cleaner and the new registers do not have, to use a technical term, bugass weird usage restrictions.

Re:Hmm

[Anonymous Coward]

The real advantage will only be indirectly felt by consumers: 64bit processors pave the way for easier, cleaner and thus cheaper software development. Remember the discussion about why the Hurd doesn't handle partitions bigger than 2GB? These are the kinds of problems which were solved with segmented address space the last time. Server applications are already running into the same problems again and some design choices for desktop operating systems and applications are already affected by a lack of address space. They work fine, but a cleaner implementation might have been chosen if 64bit processors were available now.

Re:Hmm

[X-Guy]

You are going to see a problem with 32 bit processors shortly, namely that you run out of address space. While Pentiums can support more than 4 Gig of memory, each process can only have 4 Gig of address space. This includes mapped devices like video card framebuffers. Applications use more memory and more memory each year and with framebuffers growing too this compounds the problem.

Re:Hmm

[maloi]

Ahhhh, takes me back...

Is there really much consumer (not business) application for 32-bit processors? If so, where would desktop computing benefit?

Re:Hmm

[drinkypoo]

It depends on what you're up to. Some operations which are basically limited by the number of bytes you can load into a register at a time will happen twice as fast on a 64 bit processor once optimized to use registers that size. (For some C functions and other high-level language constructs this will happen with a recompile.) The x86-64 also has twice as many registers (of twice the size) when compared to x86, so in theory you only have to move things to the stack or push them out into memory somewhere 1/4 as often. That's not really a benefit of 64-bitness (well, half of it is) but I think it's worth mentioning. I would speak with specifics but AMD still hasn't sent me my copy of the hammer specifications and programming guide.

A sibling comment to this one points out the precision issue, a very good argument I wouldn't have thought up.

But basically, anything where you're dealing with a 64 bit (or larger) value is going to get faster. Anything where you're dealing with shoveling data with a minimal amount of processing in the middle is going to get faster, too. I'm sure that a 64 bit chip with an integrated memory controller is going to have plenty of operations which let you do memory to memory copies, and you can usually only do them at a certain size, which tends to be about four times your maximum word length. This means that a single copy operation could go from 128 bits to 256 bits, but again I'd have to look up the references for intel (I have that one already) and AMD's hammer (not here yet) to doublecheck.

Re:Hmm

[Screaming Lunatic]

There's really not much need for 64bits even in gaming...but the 64-bitness of the chips is not at all important for games for the foreseeable future.

That's the biggest bunch of crap that I've ever heard. There are a bunch of games that do fixed point math because floating point does not give you enough accuracy.

Collision detection would certainly benefit from improved precision. Physics suck in games because it is difficult to do fast and accurate at the same time.

Epic has promised a 64bit version of games. I'm guessing they are doing so for a very good reason. And they are doing this despite the fact that they use a comparitively very robust physics engine in Karma.

I'm guessing you've never implemented a physics engine or even taken a Numerical Analysis course or read any books. So how about pulling your head out of your ass before disseminating FUD.

Re:Hmm

[DataPath]

I don't think its so much an issue of "does it improve performance for the consumer" but, "can we match consumer performance, and get quantity up so that it will be profitable to produce"

remember... almost all those major server processors out there were designed and are sold to a very limited number of computers... if AMD can make a server processor, and sell it to the low-end servers, business, and home computers, they can make it profitable to produce the darned thing.

At that point, home computers and servers can have mostly the same silicon, produced cheaply, and advances benefit both markets very quickly.

The way I see it, AMD just wants to merge the two markets some and pick up the economies of size.

Re:Hmm

[timeOday]

Besides, remember those pointers address *virtual* memory. I can eaily imagine wanting to mmap a DVD or big database even if I didn't have that much RAM.

Re:Hmm

[MisterFancypants]

There are a bunch of games that do fixed point math because floating point does not give you enough accuracy.

Care to name some? Didn't think so, asshat!

Re:Hmm

[MajroMax]

I would speak with specifics but AMD still hasn't sent me my copy of the hammer specifications and programming guide.

You also? I ordered mine around the middle of August, but the only thing I've heard from them since was an e-mail in October saying it was "finally available" and that they'd be shipping to my correct address.

No books.

Re:Hmm

[psamuels]

Some operations which are basically limited by the number of bytes you can load into a register at a time will happen twice as fast on a 64 bit processor

That's why God created MMX, 3DNow!, SSE and Altivec. They take operations where you are shoving lots of data into a bunch of registers in parallel and doing math on them.

Memory bandwidth is a whole separate issue, and in my opinion that's where the Hammer will really shine, in the home market at least.

64-bit registers will be nice for FEA and other HPC, the big address space will be nice for the business backend, like monster databases ... but honestly, even games do fine with the various SIMD extensions found in today's 32-bit consumer CPUs.

All of which is not to say I'm not looking forward to getting me a Hammer.... (:

Re:Hmm

[drinkypoo]

Well I'm forced to play my hole card, IPv6 addresses. As we know the size of an IPv4 address is 32 bits and a IPv6 address is 128 bits. A machine with 32 bit registers will need to use four of them to store an entire IPv6 address at once. A machine with 64 bit registers will only need to use two.

While x86-compatible CPUs have generally not been used in dedicated networking devices until very recently due to the cost to performance ratio, they have become a fairly popular high-performance embedded solution lately. Hammer should be an extremely attractive solution in the high-performance embedded space because:

• It is inexpensive. The interview that has spawned all this discussion seems to indicate that they will be in the same price range as the current 32 bit Athlon offerings. (AMD's processor roadmap indicates that only the 2-8 way SMP version (Slegehammer) will be called Opteron, and Clawhammer will be known as Athlon 64 and Mobile Athlon 64.)
• It has an integrated memory controller. Those willing to use hypertransport between their custom silicon and the CPU might need only a north bridge, or perhaps even a simpler solution. A Hypertransport to PCI bridge would be sufficient for most needs, if not all, given the use of hypertransport to do whatever actual work you have in mind.
• It is 64 bit, and as such can shovel a lot of data quickly, or work with large integers (Again quickly, since you can work with an integer of any size on a CPU with at least two digits' worth of space or at the very least one digit and a carry/borrow flag.)
• The part is (well, will be) available with very high clock rates which while it is by no means the only defining factor in the performance of a system, it is nonetheless a factor.

Another nice factor of using large integers instead of floating point is that when you absolutely positively have to get the result back in the same number of cycles each time, you can do this. Math coprocessors are just that, coprocessors. I haven't kept up so I don't know just how fast you can expect things to come back from them these days, and if they are actually scheduled or not, but at least in the olden days you had to shovel the data at the math co, then query it to find out if it was done. One problem was that if you queried it too fast it might not have set the flags properly yet and you would get bogus results. Ah, x86 is so classy!

The address space will become significant to us all very quickly if we start doing entirely memory-mapped I/O. Isn't this an issue of the Hurd at the moment? While there are other ways to solve it (but who wants to deal with segmented addressing? not me!) certainly there are many advantages to mem-mapped I/O.

And finally, sure games do fine, but more power means bigger, shiner games with more gibs! Also the reason GPUs have become so popular is that CPU speed wasn't growing fast enough to satisfy the desires of the game industry. Expect to see some more graphics-related processing to be done in the CPU for a while, namely multires (the reduction of vertices in a model one at a time with re-meshing in between, with the greatest number of vertices assigned to the appropriate models and usually determined by a scoring system, using very high-vertex-count models which may never be rendered with all visible vertices plotted EVER.) Multires and the most simple of occlusion techniques is enough to make a scalable game which will look very good on even low-end hardware and still look fantastically better on high-end equipment. It does cost you CPU though, and I'm sure you can see where I'm going with this. Of course multires will be an inherent feature of a future generation of 3D accelerators which will do even more for the developer and likely have even crappier drivers.

Also the memory bandwidth of hammer doesn't seem like it's all that outstanding except that it's integrated into the CPU and so you can expect to do less waiting. The real advantages in terms of memory bandwidth will be in SMP systems. Of course I don't know too many people planning to go to Clawhammer who aren't planning to go to dual Clawhammer, but if they are less inexpensive than promised I'll be one sucker with only one of 'em.

Re:Hmm

[Znork]

Um, games like EverQuest already touch the 1GB memory requirement for the game to run optimally. 3GB is the current max memory per process in a 32bit architecture. That's just a factor 3 off from the hardware being unable to support the game well. How long did it take to go from 256MB being plenty to 768MB being recommended?

Of course, it's to a large extent due to sucky programming, but it's not far off.

Re:Hmm

[Xsession]

what about multi-tasking? can i realistically expect to do ps7, dreamweaver, plasma, and other necessities like winamp etc. all at the same time? and i dont mean having them all open but processor intensive work. granted all these apps are 32bit will they run faster (together) on a 64bit cpu? games run fast enough there's not much scope for a 64bit gaming desktop unless you're $loaded$

Re:Hmm

[psamuels]

Well I'm forced to play my hole card, IPv6 addresses.

Ha! I made drinkypoo play his hole card. I confess I never thought of using a server-class CPU for the embedded market such as routers. Perhaps it will make sense. But that's still not to say 64 bit math is needed in the home market.

Another nice factor of using large integers instead of floating point is that when you absolutely positively have to get the result back in the same number of cycles each time, you can do this.

Uh, no, due to memory caching - and OS interrupts and such - there's very little you can depend on for absolute clock cycle numbers.

Also, you are out of date concerning the 80387. AMD recognised the shortcomings of the 8087, uh, design, so they came out with 3DNow! for a more modern approach to floating point. Intel countered with SSE. I don't know modern x86 assembly but from what I hear, both are a lot pleasanter to work with than 8087 ever was.

The address space will become significant to us all very quickly if we start doing entirely memory-mapped I/O.

That's one of those cases where "a use will be found" if the capability exists. Applications designed for huge data sets seem to have no trouble paging things in and out manually; I imagine there are layers that do this semi-automatically for some languages. A 64-bit CPU and accompanying OS will solve this problem, but it's hardly essential. Now, when the average joe starts getting computers up in the 4GB of memory range (and yes, I can see we're getting there), then it makes sense to have a CPU that can address more than that at a time..

Re:Hmm

[einer]

Not one to pick nits, but his post didn't disseminate any fud. I wasn't the least bit fearful, uncertain or doubtful after reading his post.

Misiniformation? Yes. Fud? No.

Re:Hmm

[hesiod]

Another Clippy, just what we need.

Re:Hmm

[SoftwareJanitor]

$60 for 512M of PC133? You are paying too much. Fry's is selling 256M PC133 sticks for $15 this week, and the 512M ones are only a little more than double that. But who wants PC133 these days... most of the new motherboards support DDR...

Re:Hmm

[rodgerd]

The New Zealand market isn't quite as price competitive as the US one, unfortunately.

What desktop users want to know..

[m0i]

will it be faster than 32 bit offerings? For almost anyone out there, it's the only factor when buying a CPU: speed! Adressing >4Gb of memory is not that worries me first :)

Re:What desktop users want to know..

[FearUncertaintyDoubt]

For almost anyone out there, it's the only factor when buying a CPU: speed! Adressing >4Gb of memory is not that worries me first :)

For desktops, you are right. However, a huge part of the 64-bit market is in servers, and the possibility of >4GB memory is a Big Thing. My SQL Servers will eat that much for breakfast.

Re:What desktop users want to know..

[Junks Jerzey]

For almost anyone out there, it's the only factor when buying a CPU: speed!

Nope. These days it's price. You can barely, oh so barely, tell the difference between 866MHz and 2.4GHZ, and only then when running certain high-end games or 3D modelling packages. Now go over to Dell's site and price a 2.4GHz system. You can easily get something with 256MB and no monitor for US$800. Now upgrade to a 3.06GHz P4. How much does that does that 27% increase in clockspeed cost you? Just over US$1000. And what does it get you? Remembering that clockspeed does not translate directly to more CPU performance, maybe you're getting a 20% across the board improvement, but _man_ are you paying for it, both in cost and power consumption. And was it worth it, for 27% faster than "more speed than I know what to do with?" Probably not (though I realize that all hardware site weenies will absolutely insist that they can feel the difference when browsing the web on such a machine).

Re:What desktop users want to know..

[perbu]

Memory bandwidth. A 64bit CPU can fetch and store data over a wider bus.

Re:What desktop users want to know..

[Anonymous Coward]

will it be faster than 32 bit offerings?

Yes it will, due the larger register file of x86-64. Epic ported UT2K3 to x86-64 and said they saw a 15% perf increse vs. IA32-version running on same CPU.

Re:What desktop users want to know..

[sv0f]

And what does it get you? [...] Maybe you're getting a 20% across the board improvement, but _man_ are you paying for it, both in cost and power consumption. And was it worth it, for 27% faster than "more speed than I know what to do with?" Probably not (though I realize that all hardware site weenies will absolutely insist that they can feel the difference when browsing the web on such a machine).

Tubes versus solid-state...
Beta versus VHS...
Vinyl records versus CDs...
Air-cooled versus water-cooled...

Re:What desktop users want to know..

[Anonymous Coward]

The difference between 866MHz and 2.4GHz is very apparent -- that's simply too big a difference to go unnoticed even under light load. A system starts to "feel" faster when it is about twice as fast. You won't get your work done any faster, but it is a much more reactive and enjoyable machine to work with. I agree on the pointlessness of a 27% higher core clock of course, especially when the RAM bus speed is lagging as far behind as it is in that kind of system.

Re:What desktop users want to know..

[jonnythan]

You can easily get a 2 GHz system with 256 MB RAM *and* a monitor for less than $600.

Re:What desktop users want to know..

[BlackHawk-666]

But don't bother trying to do this with Windows, since it can only handle 4GB of RAM, unless you run Datacentre edition.

Re:What desktop users want to know..

[Anonymous Coward]

Most modern buses do not fetch 32-bit words over
the bus when a 32-bit processor does a memory read.
Instead they read about a cache-line's worth of memory, taking advantage of the tendancy of our programs to reference nearby memory, nearby in time.
You can get surprising speedups by ensuring that your data is laid out in memory to take advantage of this (and the way the cacheline is filled - sequential lr versus wraparound from the referenced word's entry).

Re:What desktop users want to know..

[Decameron81]

Yes the main factor when buying a new CPU is price atm.

But you DO notice the difference between a 866 Mhz processor and a 2.4 Ghz one in many ways. On of them is the time it takes for the computer to boot. But there are several other tasks that become much faster by going up with the frequency... also remember that a 2.4 Ghz processor has DDR whereas an 866 Mhz one probably won't (haven't heard of 866s with DDR, although I may be wrong). Hopefully another factor that will show you a nice speed increase in the future is the new Hyper Threading tech in the 3.06 Ghz Intel CPU.

The computer's overall speed is increased, and yes, you will notice the big difference when it comes to playing games, using programs like Pro Tools or doing Graphics, but that doesn't mean the rest isn't changed at all.

I have a K7 850 and an Athlon 1400 DDR and hell, do I notice the difference? Of course I do...

Decameron

Re:What desktop users want to know..

[borum]

Yes it can, but usually it won't :)

Most memory accesses are cache line-refills and word-stores. The refills are often done on 64 bit busses anyway (maybe even 128b), and stores are not important latency-wise and can be grouped by a writebuffer to use the whole bus.
So the internal register size has nothing to do with memory bandwidth...

/K

Re:What desktop users want to know..

[Mitreya]

You can easily get a 2 GHz system with 256 MB RAM *and* a monitor for less than $600.

Sure and it will exactly as crippled as you might expect for such price. MB with built in crap + slow cheap harddrive + crappiest of GeForce2 series. And you'll get crappy RAM and a crappy case...
Alternatively, a good expensive CRT monitor could cost you $600 (mine is $400 but its refurbished)

Re:What desktop users want to know..

[Anonymous Coward]

The computer's overall speed is increased, and yes, you will notice the big difference when it comes to playing games, using programs like Pro Tools or doing Graphics, but that doesn't mean the rest isn't changed at all.


You misunderstand hyperthreading. You will only notice the difference when it comes to playing games while using programs like Pro Tools while doing Graphics.

Re:What desktop users want to know..

[Jeff DeMaagd]

A system starts to "feel" faster when it is about twice as fast. You won't get your work done any faster, but it is a much more reactive and enjoyable machine to work with.

If CPU performance is the _only_ differentiator, maybe. Back in the PII days, I had a 166MHz Alpha (21066) for a personal machine and at work, a PII, 300MHz, both systems ran Windows NT 4.0.

In just about every way, the PII was supposed to be two times or more times better, but the Alpha system actually felt "snappier", despite the computation performance and bandwidth of the PII being tested as 2x faster.

Re:What desktop users want to know..

[Anonymous Coward]

...or when the game does physics calculations in one thread and graphics stuff in another thread.

Re:What desktop users want to know..

[greenrom]

You may not need >4Gb of memory today, but you will probably want it sooner than you think. Back in the 80s, the barrier was 640K and that was more than enough memory to run a full-featured word processessor like Word Perfect. I doubt you could even run notepad in 640K now. Bloatware isn't going to end anytime soon. The 1GB of RAM I've got in my system is plenty for the software I run today, but in a year or two, I'm sure I'll be well over the 4GB barrier.

Re:What desktop users want to know..

[tswinzig]

You can barely, oh so barely, tell the difference between 866MHz and 2.4GHZ, and only then when running certain high-end games or 3D modelling packages.

Sorry. Wrong. I went from a 1Ghz Athlon to an 1850Mhz AthlonXP. I use Windows XP. Programs opened faster. And when you're talking about Mozilla, or Office, or Photoshop, or Dreamweaver, or anything more complicated than notepad, really, you DO notice this. Especially when you're opening and closing programs all day long.

When I come across a webpage designed with complex tables and CSS elements, the speed improvement is noticeable (e.g. my banking website, which I frequent, is complex and now renders much faster).

You can never have enough speed. You will always notice a difference, eventually, because the more power that becomes available, the more complex things become that we use frequently.

And believe it or not, but many people like to play new games. Not just "gamers." Regular people, too. My dad can barely turn around in Quake, but he loves wandering around in god mode and shooting things. He wants to play Doom3 when it comes out. He will need new hardware.

I'm just sick of this lame argument that people aren't interested in new processors because they can't tell the difference between 800Mhz and 2Ghz. Bullshit. They might be able to LIVE with the difference in speed, especially if money is tight, but you can never have "too much" speed.

Re:What desktop users want to know..

[FearUncertaintyDoubt]

The current versions of windows only run on 32-bit processors anyway. MS is releasing 64-bit versions of 2000 server sometime soon, which will be able to address memory in the TB range.

Also, you can actually run up to 8GB of RAM in Windows 2000 Advanced server.

Re:What desktop users want to know..

[CableModemSniper]

If I ever meet you I will kick your ass. No seriously. I am not joking. What a jerk. "Pah! Spending 600 dollars on a computer! What a piece of crap! Everyone must have the most bleeding edge machine ever! Also i paid a lot of money for my monitor because I am a snob. But I got a deal on it because I am soooo smart."
And just to make absolutely sure this post destroys as much of my karma as possible, I'm gonna post with my bonus. -1 Flamebait here I come.

Re:What desktop users want to know..

[CableModemSniper]

was the port just a recompile or did they tweak it? Inquiring minds want to know.

Re:What desktop users want to know..

[Anonymous Coward]

Why are you so arrogant to believe that you'll be moderated at all?

What a fucking dipshit.

It's sad, too, because I sided with you until I spotted your arrogance. You can actually build a GeForce 4-level, AMD 1.4+ghz level computer with a monitor and decent case for $600 using www.newegg.com. So, you were right. But then you were a presumptuous ass. Oh, you're soooooo fucking important that a moderator is going to give two damn bitshits about your dumb post and "destroy as much of your karma as possible."

Fucking dipshit. Go choke on a goddamned bat leg.

Re:What desktop users want to know..

[Anonymous Coward]

Life without assholes like the both of you would be very boring. Thanks.

Re:What desktop users want to know..

[Anonymous Coward]

Your analogies (I assume that's what they were -- you offerred no explanation) are flawed.

CPUs are quantitatively different only. There are no qualitative differences between 800MHz or 2GHz. It's simple, measurable performance. You measure it. It's there. Then you decide if the x% improvement is worth the $n.

But with tubes or beta or vinyl technologies... these are purely analog formats with qualitative differences in how they represent or modify analog signals. This is why audiophiles can sit around and masturbate to romantic notions of tube 'warmth' or vinyl resolution, etc. WE ARE NOT LIKE THEM. You can tell me that your 2.4GHz is better than my 800MHz, but if you're right then at least we both agree that the difference is PURELY a clock cycle one. There is no characteristic or quality that your gear has that mine does not.

Your air-cooled versus water-cooled analogy doesn't hold either. That's simply a reliability and cost issue. Your computer doesn't run "better" if it's watercooled. It runs "better" if it runs faster. So the debate is not about cooling technology; againk, it goes back to CPU speed.

Re:What desktop users want to know..

[Mitreya]

WTF, man... I am not trying to troll here. You can buy a decent computer for $600 depending on your needs, sure. But in my opinion you can at best get a decent Athlon machine AND you have to assemble it yourself (or you pay someone to assembe and maybe warranty it). P4s are just more expensive and thus something else must give. Read the discussion in tomshardware.com

As to the monitor -- I had to get a refurbished 'cause I can't afford a good new one. And any bleeding edge machine costs AT LEAST 1.5K-3K but that's beside the point.

Re:What desktop users want to know..

[Dave_bsr]

Athlon 1600, newegg.com, $56
ECS K7S5A mobo, newegg.com, $54
512MB of kingston PC133 memory, newegg.com, $50
Maxtor 80GB HD, newegg.com, $110
Liteon 48-24-48 CDRW, newegg.com, $50
Chieftec 450W Tower, newegg.com, $55
SB Augigy+firewire, newegg.com, $60
floppy, NIC, etc, newegg.com, $30
Shipping, Fedex + newegg.com, $50 (approx)


Building a sweet, powerful, linux-ready system for ~$515?
Priceless.

ps- If my math is off by a bit, sorry. And I never checked shipping, but newegg's is cheap. And, does anybody know about Turtle Beach Santa Cruz support under linux? Experience?

Re:What desktop users want to know..

[Mitreya]

Precisely! To get a decent machine so cheaply you must build your own athlon. And still you can not fit a decent monitor (15" don't count) for under $600.

Note how you built an athlon with PC133 memory. Now compare that to a P4 system (that does not work well with PC133 memory, I believe). And I assumed that original post referred to already built machine. Now you believe me that you can't get a good assembled P4 system for $600?

Re:What desktop users want to know..

[Dave_bsr]

Dave_bsr's Law: Governing "when computers will be fast enough":

Computers will be fast enough, when, for every conceivable operation, system delay between user requests and proper system response is less than the human ability to resolve, eg, instantaneous.

Not instantaneous, as in .05 seconds (button click speed now now), no, I mean _instantaneous_, like when I push on my door and I see movement. When I write on paper and _instantly_ there is writing. Then, computers will be fast enough. And I don't just mean speed for mozilla, i mean processing real-time 3d bump-mapped, whooseyourdaddy environments. Yeah.

Re:What desktop users want to know..

[jandrese]

Er, you missed the point. Below a certain price point on hardware, the corner cutting begins to effect the hardware in a seriously detrimental way. If you buy the $600 computer, you are probably going to pay more in the long run than the guy who paid $800, because he can keep his computer 4 years whereas yours is breaking down after only a year. Worse, if you try to upgrade your $600 system, you often find that there is no driver support for the crap hardware in any OS beyond what it shipped with. Also, support is generally terrible on extremly low end hardware like that, not that it helps because you will have very little warentee anyway (the people who build these things know that they break down a lot). It's certainly true that you don't need to spend $2000 on a computer to read email, but if you go for the abolute bottom barrel system, you are going to get what you pay for.

Re:What desktop users want to know..

[Dan Ost]

It's certainly true that you don't need to spend $2000 on a computer to read email, but if you go for the abolute bottom barrel system, you are going to get what you pay for.

Has anyone tried out the $199 computers from
Walmart? These have got to be the absolute bottom
barrel systems you can get. Are they crap?

Re:What desktop users want to know..

[Ninja Programmer]

• will it be faster than 32 bit offerings?

Uhhh ... yes? (Hint: it will have higher clock rate and execute more instructions per clock (on average) than Athlon)

• For almost anyone out there, it's the only factor when buying a CPU: speed! Adressing >4Gb of memory is not that worries me first:)

Well it will be faster. Look, I don't know about you, but I have 512 MB of memory in my system. I have estimated that system memory capacity roughly doubles every year. At this rate, in 3 years we will be at 4GB in mid to high end machines, just as a natural course of things.

The way I read that, in three years, 32bit-only systems will be obsolete.

Re:What desktop users want to know..

[shepd]

Bingo!

ECS == PC-Chips == Illegal, Remarked, Fake, Pirated Crap.
5400 RPM == Low end crap.
Liteon == SHIT. Yes, I used capitals and swore for a good reason.
Chieftec == ??? I've no clue who they are, but if they are using a "450 Watt" power supply that may supply up to 430 Watts of usable power if you trust the labelling (which, in and of itself is wrong -- since when does 12 * 15 = 380, and since when does 92.4 + 150 = 220?), they're a joke. Either that or this is their cheap-ass line of cases.

I've built a lot of systems, and yup, some of them have a similar configuration to yours. I build these for the customers who want to spend nothing on a system, and while they work, they are _always_ the worst PITA to make work. And they usually break. Often. K7S5A wasn't too bad, but why rely on a brand with such a pathetic reputation (don't believe me? Talk to a local retailer, or just lookup PCChips on deja or google)?

Either way, where's the monitor?

Re:What desktop users want to know..

[me3head]

Just got a Dell P4 system for $512 P4 17GHz 128MB ram 40gb hdd DVD drive Int. Sound ** Geforce 4 4400 w/ DVI (could hve had the computer for $80 less w/o this)

Re:What desktop users want to know..

[CableModemSniper]

Hey at least I had the balls to be a logged in asshole. And someone oviously gave a fucking shit about my post namely you. And you ddin't even mod me, no you went through all the fucking effort of actually writing something in response. I'd say I was pretty fucking successful in pissing someone off.

Re:What desktop users want to know..

[CableModemSniper]

Oh and btw, I love the guy who decided I was overrated. Afraid some dick in M2 is going to decide that my post wasn't really Flamebait and ruin your karma?

Re:What desktop users want to know..

[sirsnork]

Actually MS already has XP Pro running on 64bits

Re:What desktop users want to know..

[bfree]

I think the sort of latency you see from current "realtime" audio systems is about right to meet my demands, <2ms! When we can have our computer provide a simulated environment at the compexity of modern physics in which we can shoot at our friends driving F1 cars while impregentating our girlfriend, all modeled indistinguisably from real life (could we ever model the creation of a human life meaningfully). That should keep us going for a few years but it will give scientists an incredible toy!
With up to 10^81 (2^273) atoms in the universe and then what level of subatomic detail? 512bit seems about right to me. Pity Moore's Law suggests we might have to wait til around 2674 til /. will be covered with questions over whether 512bit really is enough, with processor speeds of 3 Peta9Hz (thats 3 + 9 * 000,000,000,000,000). Who would want a beowolf cluster of those? Whose going to supply the power!
Unfortuantely anyone who reads this today will be lucky to see a nice 128bit computer in 96 years time at 1.5MegaPetaHz :-(

Re:What desktop users want to know..

[Billly Gates]

It depends on what you do.

I upgraded my pentiumIII700 to an athlon XP1800 and was dissapointed initially. The performance seemed the same when launching Word or even quake3 unless quake3 was set at the maximum resolution. These apps are what most people run there computers for. IE, word, and excell is the norm and any pentiumIII or pentiumII can do the task.

However I switched from Linux to Windows2k due to constant rpm hell and noticed only after cygwin, mozilla, mysql, openoffice, netbeans, and other opensource apps were installed that I noticed a big difference. Even this newer athlon seems slugish during bootup but it was alot more bearable then my pentiumIII. So for me it matters but with most people it does not. Nobody besides developers runs these apps.

I now have a cable modem so I can finally leave redhat and suse for real distro's like gentoo or debian and believe me gentoo is noticable faster when doing an "emerge sync" or "emerge kde".

800mhz is fine for %85 of all users at there.

Re:What desktop users want to know..

[be-fan]

Actually, the original Pentium bus was 64 bit, and that has been the widest one until dual-channel DDR became all the rage. A 64-bit chip doesn't necessarily have any more bandwidth than a 32-bit chip, especially since main memory is accessed in cachelines of 32-64 bytes regardless of the width of the chip.

Re:What desktop users want to know..

[Anonymous Coward]

Hot mama! 17GHz? Sweeet

I couldn't even get a 3.06GHz HT for less than a grand (CDN).

(Yeah, yeah, I know. 1.7GHz)

Re:What desktop users want to know..

[grape_soda]

well, not walmart but last year, maximum pc got ahold of a KMart machine... lets put it this way:

it wouldnt even post/run half the benchmarks that they tried to run on it :)

Re:What desktop users want to know..

[Mitreya]

I don't know where you got it. A quick check on dell.com shows a quote of 599 for their cheapest P4 1.8 and that DOES NOT include a GeForce4 4400 (which is worth slightly less than 200 bucks). Are you a dell employee by any chance?

Re:What desktop users want to know..

[grape_soda]

have you seen the reviews on the new P4? that hyper threading needs some work. from the reviews ive read the hyper threading provides MAYBE a 4% increase..

Re:What desktop users want to know..

[grape_soda]

surely if we can make a computer with that kinda power, shouldnt we just be able to transplant our brains into some "liquid ooze" and merge with the machine? i mean c'mon.. how fast does our brain compute? faster then 1.5MegaPetaHz?

Re:What desktop users want to know..

[teamhasnoi]

Thank you - I have been saying that for a long time, you have just said it better. :)

I say that to people and they look at me like I'm the guy who microwaves a burrito for 30 seconds and drums his fingers while growling, "Damnit! Hurry up!" There is no reason we shouldn't be closer to that than we are now; bloated code will hog a processor no matter what the speed. We could probably have near instant results with what we've got right now - if the programs weren't so big Bertha.

I'm starting to think that Bill is compensating for something.

Re:What desktop users want to know..

[Anonymous Coward]

Hmmm.. power to crack RSA-8192 or AES-256 instantaneusly... now that's what I'd call enough ;)

Re:What desktop users want to know..

[cerberusti]

just fyi from windows xp task manager, notepad.exe, immediately after opening takes 2,832k... kind of puts things into perspective.

Re:What desktop users want to know..

[Anonymous Coward]

Sorry. Wrong. I went from a 1Ghz Athlon to an 1850Mhz AthlonXP. I use Windows XP. Programs opened faster.

I bet you re-installed your Windows. Fresh Windows is allways running better, since most of Windowses do have degrading performance in desktop use. Registry is for one. Don't know much about Windows XP, but at least NT 4.0 and Windows 9.x does this.

Re:What desktop users want to know..

[TheLink]

Yes. Otherwise AMD will get killed.

Not faster = mistake. AMD cannot afford mistakes.

Intel can survive launching a next gen processor that is slower than current. But AMD can't.

Technical reasons for faster/slower aren't that relevant in this case. AMD has to execute or be executed.

Re:What desktop users want to know..

[Decameron81]

Well to be honest I will wait to see some user reviews to decide wether it's crap or not. But to be honest I like the "concept" behind it, since Hyper Threading may represent a small speed gain atm, but this small speed gain will grow with faster processors.

A small gain today may become a big difference tomorrow.

Decameron

Re:What desktop users want to know..

[fredrik70]

I agree with you to a point, but I must say I do find there's is a noticable difference between a 866 and s 2.4G machine. Maybe not for everyday tasks, but if you got a compilation running in the background you be happy for that extra horsepower.
But yes, buying the latest and greatest of CPUs might not be the cleverest thing to do from a price/performance perspective.

Re:What desktop users want to know..

[fredrik70]

Not that I'm a doctor but IIRC our brain is very, very slow compared to any CPU of today, it removes that weakness by working in parallell, very much so as well.

Re:What desktop users want to know..

[Xsession]

i've seen reviews that indicate the hyper-threading decreases performances :(

Re:What desktop users want to know..

[Xsession]

the futures lies in quantum multi-processing! 5000 processors in your bedroom under your bed running off your body heat :), still taking 2 mins to start up windows :(

Re:What desktop users want to know..

[Anonymous Coward]

If your SQL Servers eat 4G for breakfast, you need a DBA... or a reinstall.

Re:What desktop users want to know..

[pmz]

You can barely, oh so barely, tell the difference between 866MHz and 2.4GHZ, and only then when running certain high-end games or 3D modelling packages.

This is very true. I've been using Pentium-class PCs and UltraSPARC II machines for years. Recently, I got to try out a friend's 2GHz Pentium 4 Dell.

The result: the hard drive is so slow that it is essentially the limiting factor in the P4 system (it was a 7200RPM drive, too). For general use, I really don't see much practical difference among all the computers I use and the P4 system. I will admit that the P4 did very well with 3D games, and I'm sure things like Flash-based web pages will do well, too. Thankfully, I have a Playstation for games and don't care about Flash.

The need for faster CPUs and more memory can be determined only by an individual's true need. 3D modeling and scientific systems should get all the power available. General-purpose desktops maxed-out quite a long time ago.

Re:What desktop users want to know..

[tswinzig]

I bet you re-installed your Windows.

Uhh no, why the hell would I do that? All I did was change my CPU.

Fresh Windows is allways running better, since most of Windowses do have degrading performance in desktop use. Registry is for one. Don't know much about Windows XP, but at least NT 4.0 and Windows 9.x does this.

You might want to look into something called 'disk defrag.' Uhhhhhhhh

Re:What desktop users want to know..

[FearUncertaintyDoubt]

I am a DBA. The ideal case for a SQL server is to be able to load then entire database into memory (or most of it, anyway). Unless you think that no databases exist over 4GB...

Re:What desktop users want to know..

[NeMon'ess]

I hope you're planning on buying more RAM from somewhere else. You should have 512MB to effectively use the 4400. With less than 512MB you're going to have the HD slowing down performance massively.

Re:What desktop users want to know..

[BlackHawk-666]

To clarify things:

Windows 2000 Pro - 4GB Windows 2000 Server - 4 GB Windows 2000 Advanced - 8 GB Windows 2000 Data Centre - Lots!

Wow

[r0xah]

Yes 64 bit CPU's for desktops will soon be the next new thing, but who really needs them? Grandma and grampa checking their email won't need something that fast and even the normal computer user will never experience such CPU intensive work to need a larger word size. Trust me I am not saying I won't be one of the first people to run out and get one, but there really is no need for the general public to have 64 bit processors.

Re:Wow

[scotch]

Grandma and grampa checking their email won't need something that fast ....

Grandma and grandpa could check their email on a 16-bit computer. Don't forget grandpa's geri-porn, you need some horsepower for that.

Re:Wow

[jki]

. Trust me I am not saying I won't be one of the first people to run out and get one, but there really is no need for the general public to have 64 bit processors

Now, when you are a rich and famous IT star, you will regret saying that like this guy does :)

Microsoft has not changed any of its plans for Windows. It is obvious that we will not include things like threads and preemptive multitasking in Windows. By the time we added that, you would have OS/2.
-- Bill Gates, from "OS/2 Notebook", Microsoft Press, (c) 1990

Re:Wow

[dingleberrie]

It's not about what consumer needs 64 bit for today's applications... it's tomorrow's applications. First there must be a base of users out there.

Do you remember the opportunity brought about by the 386? Who needed that when all the modern applications ran fine with the 286? The 386 even broke some of the old 286 code. But it was still very useful to programmers who could spend focusing on quality (and bloat?) rather than worrying about how to confine data to 64 K blocks. Almost 20 years later we are still benefitting from the whole flat memory model that finally came to x86 (flat up to 4 GB, that is).

If you have to ask the question of who needs it, then it's not you... yet. Sure the first adopters are the Corporate people who know they need it as well as the "look what I have" crowd. But I'm pretty sure that there will be consumer applications that will make 64 bits necessary after there is enough consumers that have them.

640 TB should be enough for anybody.

Re:Wow

[Anonymous Coward]

Yeah, well, grandma and grandpa will be dead soon, so what then??

Re:Wow

[Anonvmous Coward]

"Grandma and grampa checking their email won't need something that fast and even the normal computer user will never experience such CPU intensive work to need a larger word size."

That's a bit of a narrow minded view, don'tcha think? Consider this: We don't know what we'll be doing with computers 2-3 years from now. If it turns out that PVRs are a killer App, for example, then suddenly 64-bit processors are interesting.

The "who really needs it for the most basic stuff?" argument is extremely tired. Lots of people buy their machines based on their potential, not what they can do with them today. Don't believe me? Then look at all the people who bought an XBOX solely because of it's chipset and hard drive. They were (and are today) expecting to eventually buy games that blow them away.

If computers were strictly used for their most basic features (internet browsing, email, etc...) then 'internet appliances' would have been some sort of hit as opposed to the flop that they are. So please, put this 'how do I get my grandma to buy one?' argument to rest. The answer is she won't. But there is still a large market of people who do want/need 64-bit processing. You don't need for grandma to want one in order for the product to be a success.

Re:Wow

[RatBastard]

IBM didn't think anyone needed 386 systems. Compaq decided that they'd sell them instead. IBM's control of the PC died that day.

Re:Wow

[Anonymous Coward]

"but there really is no need for the general public to have 64 bit processors."

Gee, I seem to remember hearing quotes like "no computer should need more than 64Kb of RAM" and "there will only be 4 computers in the world".

Get with the times, technology changes (some would say "improves") every day. CPU speeds double nearly every 18 months, HD size at about the same pace.

64bit computers aren't new. I've got one under my desk at home that is at least 5 years old (DEC AlphaStation 200 4/233). The only thing "new" about 64bit computing today is that Intel legally holds the license to the architecture they've been stealing for quite awhile, and that Microsoft is back in the game.

RC3 of Windows NT 5 (later known as Windows 2000) had a 64bit version that ran like a champ on Digital's Alpha processors. But Compaq $crewed the Alpha line and licensed the architecture to Intel instead of pushing the lawsuit that DEC had against them.

Re:Wow

[TheAncientHacker]

Actually, IBM was pretty damn sure that people needed 80386 systems. What they were also just as sure about was that an 80386 based PC would canibalize sales from their System/36 systems. The folks up in Rochester, Minnesota (where the System/36 and later AS/400 come from) went to Armonk (IBM Headquarters) and had the IBM Executive Committee block the 80386 based PC.

The industry stalled for a while because NOBODY had introduced anything for the PC compatible industry that wasn't a clone of IBM's systems or peripherals until then. Finally, Compaq risked the company with the DeskPro 386 and IBM was in serious trouble.

Re:Wow

[RatBastard]

Who needed more than 64K of RAM?
Who needed MoDems faster than 1200 baud?
Who needed processors faster than 1MHz?
Who needed more than four colors?
Who needed hard drives?
Who needed mice?

I could stay with my old ColorComputer 2 but I don't want to.

Re:Wow

[ToasterTester]

You are correct sir!

Even business isn't going to have that much need for 64-bits. Very few people have a need for 64-bits of precision in doing calculations, scientific applications mainly. Then very large database could use the extra memory addressing. But we're talking the databases John Ashcroft wants to build to track every person in the U.S.

People don't have a need for a 64_bit CPU. I believe it was here are /. last week on how computer sales are down because people are realizing these new faster CPU are buying them much. Business and home users are just updating some componets in their systems and not going for the latest 2.n barn burner system. Only gamers and high end multi-media artists are going for the faster and faster systems.

AMD better look for a side business their 64-bit isn't going to keep them afloat. Because even with the 64-bit systems, you need software that take advantage of 64-bit.

Re:Wow

[Anonymous Coward]

Oh, but you ought to! Here! *Wham-wham-wham*. Feeling better? Schlick-schtick! Oompaboompa-wroom-wroom-wroom! Would you be a good girl and go get your CoCo from the basement to see if you can hammer another ounce of performance out of it? See, told you. =) RUN Oh no, it lives! Errh, perhaps just leave it alone, maybe, will you.

Re:Wow

[Arthur Dent]

Yes 64 bit CPU's for desktops will soon be the next new thing, but who really needs them? Grandma and grampa checking their email won't need something that fast and even the normal computer user will never experience such CPU intensive work to need a larger word size.

You're forgetting something: What if Grandpa and Grandma want to view that shiny video email of their grandkids? And what if they want to play movie director in their copious free time and compose a video email themselves?

After all, today's crop of digital cameras already record mpg clips (about six seconds worth before the CF card fills up), but it won't be long before flash ram gets even cheaper and we start seeing 4/8 GB cards.

Once the processors are available, applications will be written to take advantage of the larger word sizes. There's no way to tell what will happen.

Re:Wow

[Dan Ost]

Because even with the 64-bit systems, you need software that take advantage of 64-bit.

If you've got source and an appropriate compiler,
you have 64-bit software. Sure, the source isn't
necessarily any different from what you compiled
on your 32-bit machine, but it might run
significantly better on a 64-bit machine if it
uses large structs can take advantage of
parallel instructions (2 32-bit operations on
the same processor in the same clock-tick).

The masses might not care, but this is definately
a good thing.

Re:Wow

[Anonymous Coward]

Once the processors are available, applications will be written to take advantage of the larger word sizes. There's no way to tell what will happen.

Well that's easier said than done.

When you look at the 16-to-32 bit transition, the lack of general purpose software really retarded the market. People were perfectly happy with their DOS, Win16, and OS/2-16 apps and their extenders and memory managers and so on.

1987: i386 introduced.
1993: + 6 years: First fully 32-bit mainstream OS introduced (Windows NT)
1995: + 8 years: Mainstream software mostly shipping as 32-bit (Windows 95)
2001: + 14 years!!: Full 32-bit OS shipping on most PCs (Windows XP)

Especially with the economy in the toliet, let me predict a similar fate for the 64-bit desktop.

Re:Wow

[Anonymous Coward]

Yeah, I'm sure Bill Gates really regrets the millions of dollars IBM paid him to push that piece of crap while he was secretly paying Dave Culter to add threads and pmt to Windows.

Re:Wow

[Anonymous Coward]

I completely agree with you.

I have a duron 1300mhz and it is so slow to use it as a PVR. After recording a 100 min movie in MPeg2, it takes 2 hours to compress it to divx in order to burn it on a CDR.

There are many reasons why people would want, in a year or two, to use their computer as a VCR and there are many resons why people would be interested by the extra processing power of a faster computer.

For example PVRs allow you to skip all the ads by recording the movie and starting to watch it fifteen minutes after you started recording. This is something you can never do with a regular VCR (you can not play a tape and continue recording the end of the movie at the same time) but it is something people would possibly enjoy.

PVRs allow you to make as many copies of the movie as you want, save all the shows you ever enjoyed because CDs are very cheap and do not take much space.

But speed is a big problem. Today no computer is fast enough to compress a MPEG2 file in divx in a
reasonable period of time. mp3s takes seconds to compress, a movie takes 2 hours.

The fact that it is not a simple task to compress movies so people will not bother with it, is not relevant as technology always tend to become simpler. Recording mp3s, which was fairly complicated 10 years ago, is now dead easy for example, it just requires to point and click, there is no reason it should be any different with divx movies.

So we need 64 bit and the whole lot, because the needs of tommorrow are not the needs of today, just as the needs of today are not the needs of yesterday.

Re:Wow

[Anonymous Coward]

Yeah, there is. Immersive gaming needs intesive graphic/physics/A.I/soundprocessing. Speed/Precision is never enough. Companies won't try to take advantage of 64bit CPU's if there is no market... so its kind a 'atitude' to keep updated with technology for users.... Not to mention that computer isn't more the stand-alone altar for gaming/coding nowadays... integration with mp3 players, digital camera, frigde, security system... there is never enough speed.. at least now. :P

Re:Wow

[Anonymous Coward]

"as well as the "look what I have" crowd"

[...and there is me waving frantically in the background]

I can see one use...

[djupedal]

Desktop Apple's that laugh at the /. effect :)

Re:I can see one use...

[Anonymous Coward]

I can about see this... if Gates bails them out again.

Heat.

[Anonymous Coward]

Will the processors run cooler than the current 32 bit offerings from AMD?

As much as I love AMD, my box is far too loud, and I'm too damned cheap to shell out another $100 for decently quiet fans.

Re:Heat.

[bninja_penguin]

I think you might want to check your power supply if you are running hot. I have a 1700+ that stays at 104 degrees(f) all the time. Even after 8 hours of fragfests, it only goes up maybe five degrees. I am running a stock OEM CPU fan, and 350W P/S, with one case fan that came with the case. It is nice and quiet.

Re:Heat.

[Jim Norton]

Will the processors run cooler than the current 32 bit offerings from AMD?

I don't think anyone has a definitive answer for that question. However, you have to remember that the Athlon is an older part which is nearing the end of its life... Intel faced the same situation with the Pentium III beyond 1 GHz.

Silicon-On-Insulator (SOI) technology, which will debut with Opteron/Clawhammer, is supposed to reduce heat by around 15%

Re:Heat.

[Anonymous Coward]

You must be doing something wrong to have a computer that's "way too loud". Get an Alpha PAL8045 and a slow-moving 80mm fan and you'll run under 50c with no problem.

The people with screaming Delta fans don't know what they're doing.

Re:Heat.

[ottffssent]

Your box is too loud? So buy a P4 - that should give you less heat to dissipate. Oh, wait.

Buy the quiet HSF or suffer in !silence.

Re:Heat.

[RzUpAnmsCwrds]

That's what I have. 8045 + Panaflo 29CFM. Now if only my GeForce4 were civil.

Re:Heat.

[Anonymous Coward]

I recently bought an Athlon XP 1600+ with one of these for cooling, and I've never seen the temperature go above 37C, even with no case fans. For $8 + $5 shipping, that's pretty damn good. It's also incredibly quiet.

Re:Heat.

[Anonymous Coward]

"As much as I love AMD, my box is far too loud"
Isn't that more of a fan issue?

"$100 for decently quiet fans"
Where the hell have you been shopping?

You can get a nice copper heatpipe for under $30.00 US

Re:Heat.

[Anonymous Coward]

give me a link to that 15% figure, id like to see where you got that.

Big Bets on Table

[4of12]

Both Intel and AMD have been betting big on 64 bit computing and it will be interesting to see how this plays out.

Itanium 1 was a flop. Itanium 2 has respectable performance, but is not IA-32 backward compatible, where AMD x86-64 is backward compatible.

I will bet that backward compatibility will tilt the balance to Opteron and that Intel will scramble to introduce a new chip Yamhill(?) designed to provide the backward compatibility that IA64 lacks.

Re:Big Bets on Table

[djtech]

I agree on this will be an interesting battle of the CPU gods. Itanium lacking the X86 compatibility was a huge mistake. Almost as big as the Rambus SNAFU. I have not even seen a single Itanium sale anywhere out of all the new servers I see.

I plan on buying the clawhammer shortly after it comes out. I've been holding off upgrading until AMD's 64-bit CPU comes out. Why you ask? Well a much lower MHz Clawhammer will outperform high MHz P4's running 32-bit apps! Obvisouly it will be some time before 64-bit apps are out.

Re:Big Bets on Table

[nelsonal]

The only itanium2's avalable are hp's offerings (I think its the rp2600 and rp5670), they have a two way and four way that should have been shipping since August. Considering that HP is the co-developer and only one selling them, might tell you something. Although there have been rumblings that IBM and Dell will be adopting the architecture. I think Dell is waiting for a better chipset from someone besides HP. It seems like AMD is trying hard to move into the low end server business with these, but planning to sell a cut rate model to consumers to help cover their development costs.

Re:Big Bets on Table

[MagPulse]

The Itanium 2 has silicon dedicated to x86 emulation, which has been redesigned to be faster than the original Itanium. If this is fast enough, then it shouldn't be an advantage for AMD right?

Once AMD and Intel have 64-bit processors that are affordable and faster than their 32-bit products, I imagine apps will be optimized for both x86 and IA64 architectures. This could be by using separate binaries compiled for each, or just by writing for Java or .NET and JIT'ing to the host architecture. At this point, x86 emulation will only be used for legacy apps, so it doesn't have to be as fast as IA64 code.

I'm not sure how the emulation works though. Does the CPU have to switch modes using a lengthy switching time, or does the emulator just pick up x86 instructions and translate them to IA64 instructions?

Re:Big Bets on Table

[cheezedawg]

Intel will scramble to introduce a new chip Yamhill(?) designed to provide the backward compatibility that IA64 lacks.

I think you can rest assured that Intel already has a backup plan for a 64 bit desktop chip in case AMD finds a market there - it will not be much of a scramble. For now they are predicting that the P4 clock speeds will have increased enough by the time the first Hammer chips are introduced (still around 5 months) that they will outperform Hammer anyway.

And IA64 does provide backward compatibility with an on-chip emulator. It might not be the fastest right now, but it does have a lot of potential.

Re:Big Bets on Table

[fastpathguru]

"They" are retards. Intel itself says 3.4GHz by the end of 2003, barely over what they "sell" now. (I say "sell" 'cause even though they "released" it, you can't get one.) Newsflash: Prescott (i.e. .9) delayed to "Q4". (Read: "2004") Newsflash2: P4 3.06GHz consumes >100W peak. Rediculous. Bumping clockspeed isn't going to be easy there... BTW: How is P4 going to address the 64-bit issue without a major redesign? Hammer is a much more efficient chip. Uses SOI. Uses onboard memory controller. Multiple chip-level interconnects for glueless multiprocessing. More GPRs. 64-bit. Inexpensive platform. Did I mention glueless multiprocessing? I.e. 8-way becomes a PCB-design issue, not a chipset issue. Hammer is much more than mere "fast x86." fpg

Re:Big Bets on Table

[Jace of Fuse!]

or does the emulator just pick up x86 instructions and translate them to IA64 instructions?

As I understand it, AMD's 64-Bit processors actually have hardware for supporting the previous 32-Bit instructions. I could be misunderstanding, but if I'm not this will naturally mean that with 32-Bit instructions the AMD chip will outperform Intel's emulation.

Intel is banking heavily on people finally ditching x86 for good. There are good reasons for people to ditch x86, but there is one good reason to keep it: Legacy Support. How important that is will depend on the person and their needs.

Re:Big Bets on Table

[Anonymous Coward]

Money.. There's literally trillions of dollars in software floating around that people have paid for/own/use daily. Why wouldn't people want to continue to use their software on thier new chips, only upgrading packages here and there as the need demanded?

I don't think anyone is going to rush into a system if they have to replace all of their hardware AND software to get running.

Re:Big Bets on Table

[Ninja Programmer]

• Both Intel and AMD have been betting big on 64 bit computing and it will be interesting to see how this plays out.

They had nowhere else to go. If we start hitting the 4GB, and there is no solution, software developers and end-users will eventually be crying bloody murder like they were when Intel's 640KB limitation was hit. (That time Intel was slow to react -- this time around AMD and Intel are trying to have a solution in place *before* it becomes a problem.)

• Itanium 1 was a flop. Itanium 2 has respectable performance, but is not IA-32 backward compatible, where AMD x86-64 is backward compatible.

Well I like to dig on Intel as much as the next guy, but technically speaking, IA64 is backward compatible with IA32 (it does have a bona fide IA32 mode.) But its slow as molasses (they might as well be emulating IA-32.)

That being said, I don't think Windows device drivers are going to work on IA-64 (the IA-32 mode is not involved in the boot process in any way.) IA-64's compatibility is in fact a "joke", though technically there.

The backward compatibility mode in Hammer, is very different. You can boot 32-bit windows on it, play your old DOS games on it or whatever and you will not know the difference (except it will be a lot faster.)

Re:Big Bets on Table

[cheezedawg]

Intel itself says 3.4GHz by the end of 2003

They also said 3.06 GHz in 1H03, but it has been released for a month now, so they are well ahead of their posted schedule.

I say "sell" 'cause even though they "released" it, you can't get one

Not true- a quick look at several stores listed on pricewatch, and they all have the chip in stock. Dell also has been shipping systems with the chip for a few weeks now. Now stand back and compare that with the Athlon 2700 and 2800 (both announced waaaay back in September to great fan-fare on /.). The 2700 is hard to find, and the 2800 isn't even available at all.

Prescott (i.e. .9) delayed to "Q4".

While we are talking about delays, look how far back AMD has pushed Hammer over the past year. They first wanted to release it this year, now its mid 2003.

P4 3.06GHz consumes >100W peak

Not true- its peak is just over 81W. And the AMD 2700 at a much lower clock rate consumes around 75W.

How is P4 going to address the 64-bit issue without a major redesign?

What? Intel has said that they don't think there is a "64-bit issue" to address for desktop systems in the near future. I do think that Intel has thought it through, though, and they do have a backup plan in case there actually is a decent market for 64 bit desktop chips. Whether that is a major redesign of the P4 or something completely different is anybodys guess.

Hammer is a much more efficient chip.

Much more efficient with its 100 million transistors (compared to 54 million transistors on the P4)?

Uses onboard memory controller.

Sure you can save a few cycles for memory accesses, but then you have to buy a whole new processor to upgrade to the newest memory technology in 6 months. There are advantages, but there are also disadvantages.

Did I mention glueless multiprocessing?

If you like multiprocessing, take a look at HT.

Hammer is much more than mere "fast x86."

At this point, that is all that it is. There are few desktop applications right now that require the higher memory addressing and other features of Hammer. The AMD model number of the first Hammer will probably be in the mid 3000's, meaning they are targeting the 3.4 or 3.5 GHz P4 as direct competition. If Intel has reached or beat these speeds by the time Hammer is released, then Intel could be equal or better than Hammer in the only benchmark that counts (how fast does it do what I want it to do), and they can do it with 32 bits.

Re:Big Bets on Table

[BadlandZ]

Itanium 1 was a flop

1 flop? I'd have thought at least a mega-flop. So AMD will be a Giga-flop?

You know what, I don't give a rats-behind about x86 backwards compatability, I care about GOOD hardware, and MORE IMPORTANTLY, a good compiler that works on THAT SPECIFIC HARDWARE.

Where is a benchmark I can understand? One that takes the best hardware, with a standard aplication (COMPILED NATIVELY, WITH a GOOD compiler) and gives me a number I can understand before I buy it.

Say what you will, but G4 is at least GCC compiled (know the compiler) and benchmarkable... I'm still trying to figure out what compilers best for NORMAL x86. One's good on AMD's, another on Intel, and then they are OS dependant.....

How can soo many people care soo much about hardware that they will pay an extra 50% for hardware that is ONLY 20% FASTER??? All while the right compiler can make all of your apps faster by 10% to 30%, without additional hardware costs?

Come on, new hardware is great, but let's not put the cart before the horse AGAIN. Avoid the hype, ask the right questions.... Where is the good compiler for the new CPU? What are the SPECs (or other benchmarks)?

Re:Big Bets on Table

[rweir]

Legacy Support.

And this is somewhere that Linux (and the *BSDs) can shine. I can get (modifiable) source code for nearly every app on my machine. My OS (Debian) runs on several 64-bit architectures already. How hard do you think it will be to port it to 'pure' x86-64? I imagine there will be issues, but nearly all the code should be 64-bit compatible already. Oh, and if it's not, I have the GODDAMN SOURCE.

Re:Big Bets on Table

[Ryan Amos]

Opteron will probably be the victor here, but not solely because of IA-32 backwards compatability. The vast majority of the desktop market runs Windows, and you'd be stupid to think that Intel wouldn't give Microsoft a very good emulator to include in their OS. x86-64 will win out because Itanium 2 is extremely power-hungry and very expensive. It really wasn't designed to be a desktop chip. Intel designed the Itanium to compete with Sun and IBM, not AMD. That's what will win out, with backwards compatability being a secondary consideration.

Re:Big Bets on Table

[MajroMax]

Intel is banking heavily on people finally ditching x86 for good. There are good reasons for people to ditch x86, but there is one good reason to keep it: Legacy Support. How important that is will depend on the person and their needs.

Ironically, 64-bit processors are also the best time in the next 5 years to ditch the worst of the cruft with the 8086. I seem to remember, in fact, that the Opteron is doing exactly that --- IIRC, some of the legacy operations are simply not supported while the processor is in 'pure' 64-bit mode.

Re:Big Bets on Table

[Dun Malg]

Hammer is a much more efficient chip.

Much more efficient with its 100 million transistors (compared to 54 million transistors on the P4)?

Yes. Efficiency isn't about transistor count. I have a halogen light bulb with NO transistors and it draws 200W. Opteron is more efficient even with more transistors because its Silicon-On-Insulator design allows it to run at lower voltage.

Re:Big Bets on Table

[zozzi]

Itanium 1 wasn't a flop. It was designed to be a development platform, and that's precisely it's function. Intel has already said it it has 5 development teams working on the next iterations of Itanium.

Many instructions have hints in them to tell the CPU some details so that it can optimize them, but for the moment they're ignored.Still they can implement it when they want so there's lots of future potential. Also Itanium IS binary ia-32 backward compatible (eg see Vol 3, PAGE ONE: "A key feature of the IA-64 architecture is IA-32 instruction set compatibility"!) but unlike amd, it's meant as an add-on not an x86+extensions like amd. Heck even the BIOS is out with IA64, as well as IN/OUT instructions - yep they're out too.

The reports you've seen on the web are mostly flawed since :

They are running ia32 code which performs poorly (so what!)

The ia64 code they were using was based on output from not-yet-optimised compilers

Running on Itanium 1

Few real world examples: With itanium, passing parameters from one function to another is via registers not slow stack access. Typical tests would do some loops and call few functions, in the OOP world with many function calls it would fly!

There's an Itanium 1 sitting here on my desk and yes it's SLOW (feels like a 500mhz if not slower) even when booting native code. I'll test itanium 2 when HP finally decides to ship me the missing graphics card!

Therefore all I can say is: give it time! It's the software which needs most tweaking not hardware. Compilers have quite a lot more work to do and are thus not easy to port

Re:Big Bets on Table

[Anonymous Coward]

Actually, they got the emulator from DEC :)

Bug deal

[Anonymous Coward]

What about the Commodore 64? And when whas this? 1992? Sheez.

Re:Bug deal

[dubbreak]

from previous: "What about the Commodore 64? And when whas this? 1992? Sheez." ah commodore 64 = 64kb ram, not a 64bit processor.. oh that was a joke, i get it haha! To my knowledge the processor was an 8 bit.. bit of a difference.. oh hahaha pun pun joke joke..

Re:Bug deal

[Bedouin X]

Let's also not forget the fact that the C64 was closer to 1982 than 1992.

Re:Bug deal

[Pyrometer]

What about the Commodore 64? And when whas this? 1992? Sheez.

Hey what about the Commodore 128? ... now that would be twice as fast right ;)
Looks like AMD and Intel have a long way to go if it has taken them this long to get to where the Commodore was so many years ago ;)

Re:Bug deal

[Anonymous Coward]

Are you retards!?! The "64" and "128" for RAM, not CPU. The C64/128 used an 8 bit CPU (I think a Motorola 6510).

They don't *WANT* to make money?!?!

[freeze128]

Didn't AMD announce that they were no longer going to compete for the desktop CPU? And now they say that there *IS* a market for the 64 bit CPU on the desktops!!! Well, are they, or are they not competing then?

I confused!

Re:They don't *WANT* to make money?!?!

[DjMd]

I love that everyone read that story and thought it ment that they were leaving the desktop market, when it really said that they were going to diversify outside of the desktop market, as in do more in addition to their desktop market...

(a quote from first paragraph of the Forbes article "[a] strategy of developing processors for a wider range of products outside computers ...")

Re:They don't *WANT* to make money?!?!

[spoco2]

I think the problem was that no-one actually read the story... just the sensationlist Slashdot article.

All the article said was that AMD saw the ridiculous waste of time in simply jacking up the speed of processes continually... We're up to 3GHz now... and what actually requires that? Not much... so why not spend the time building COOLER chips that can be cooled in a QUIETER way... in fact, why not ship your chips with a QUIET fan, like really QUIET (why am I shouting the word QUIET? Oh yeah, so I can be heard over my AMD with noisy FAN!)...

Cooler... damn that would be nice... my media server, sitting in my entertainment cabinet... pumps out a lot of heat... it's ridiculous really... I got a relatively lowly Duron 1GHz and it's pouring the heat out.

Surely, now that they're up at 3GHz... rather than screaming towards 4GHz like mad things, why don't they work on making the 2GHz and lower cooler?

Re:They don't *WANT* to make money?!?!

[Anonymous Coward]

We're up to 3GHz now... and what actually requires that?

Moore's law requires it. Can't be a "law" without it.

My joystick has a 16 bit cpu and a cooling fan--more processing power than what sent man to the moon! Do I need it? No. Do I want it? Fuck yeah I do.

Re:They don't *WANT* to make money?!?!

[Anonymous Coward]

No actually like 20 people read that story. The rest of those assholes just post dumb shit. Like this.

Re:They don't *WANT* to make money?!?!

[uberjon]

you actually read the stories on slashdot

Re:They don't *WANT* to make money?!?!

[sickboy_macosX]

I think what they said was, they were going to stop producing x86 CISC processors that compete with intel. Within the next 5 years, I expect the Instruction Set to get so bloated for Intel x86 Chips, that RISC will come back in to play, and I think that is where AMD is looking to make their money, is in RISC. I belive that when Apple announces new generation processors, that they will be AMD, or some company of the like, making RISC processors, This is just my opinion. Instruction Sets for Intel Chips are bloated already, so I expect this to come down to a CISC vs. RISC not an AMD vs. Intel or Apple vs. Intel war, just easier, computing with power to boot. -Shon

Gamers will benifet

[Anonymous Coward]

Look at the Powerful Game Cube and PS2 with 64 bit processors, they kick the Xcrap in to obilivion! Imagine, 64 bit agp will provide wonders for games, imagine affordable QUXGA graphics at 60 FPS! Its all possible

Re:Gamers will benifet

[Puu]

Have we forgot our medication again, mmm?

Sure sure, it's all most possible, indeed! Easy now...

Benchmark's

[Anonymous Coward]

Here are some benchmarks for a Operton.

http://www.aceshardware.com/

Re:Benchmark's

[mckayc]

Shit, I was hoping for some Opteron benchmarks.

64 bits=$8=8 bytes etc???

[Devil's BSD]

OK people, I know some of you are trying to be humorous, but really the 64 bits is the size of the registers and how much data the processor handles at once. Which means at 64 bits, the processor can process (theoretically) twice as much data per second than a 32 bit processor. Which also means it can handle any number up to 2^64.

Re:64 bits=$8=8 bytes etc???

[Visigothe]

This isn't totally correct.

"64bit" refers to the size of the instruction word, not "how much data the processor handles at once". That is a function of pipelining, ALUs, branch prediction, etc. This can be proved by a recompile of a 32bit application with 64bit flags. The application won't be "magically" twice as fast.

There is something else... a 64bit app may even be *slower* as the cache can only hold half the number of words, given an equal cache size. Cache misses are a huge performance hit these days, as RAM is much slower than Cache RAM.

Of course the big difference between AMD and IBM is that the new 64bit PPC970 doesn't take a performance hit switching between 32 and 64bit applications. This has more to do with the PPC ISA than anything in the processor.

The only thing that 64bits will give "normal" users is the ability to address a *huge* amount of LOGICAL memory. In most cases, it doesn't make sense to make 64bit versions of applications, due to the above cache issue. Also, note the allusion that users will require more RAM for 64bit applications, as it will be needed to store the larger word size.

.

Re:64 bits=$8=8 bytes etc???

[Eric Smith]

"64bit" refers to the size of the instruction word, not "how much data the processor handles at once".

That's certainly not true in the x86 world. x86 instructions are variable length, and can be as short as one byte. Surely you wouldn't call the Pentium IV an 8-bit processor?

There is something else... a 64bit app may even be *slower* as the cache can only hold half the number of words, given an equal cache size.

In 64-bit mode, the only things that MUST be 64 bits wide are pointers. Instructions are generally the same size (or one byte longer). 64-bit code can still use 8-, 16-, and 32-bit data, so it doesn't force your data to be 64 bits wide unless you specifically want it. So in general I wouldn't expect too much of a degradation on cache hit performance on typical applications.

the big difference between AMD and IBM is that the new 64bit PPC970 doesn't take a performance hit switching between 32 and 64bit applications.

x86-64 doesn't appear to have any serious performance hit there either. I suspect this was a design requirement for the architecture.

Re:64 bits=$8=8 bytes etc???

[Door-opening Fascist]

There is something else... a 64bit app may even be *slower* as the cache can only hold half the number of words, given an equal cache size. Cache misses are a huge performance hit these days, as RAM is much slower than Cache RAM.

Not true. Not only do cache segments not have to be 64-bits apart, most 64-bit CPUs have at least 512kB of cache, and some (i.e. PA-RISC, POWER, UltraSPARC) can have up to 16MB of cache. Not even Intel's Xeon or Itanium can match that.

Intel

[jlechem]

well I think this will be quite interesting to see who wins the war now. I agree that the average user doesn't need 64 bits of CPU power. But then again some one was quoted as saying a person will never need more than 640K of ram either ;)

I think AMD is doing the right thing by focusing on the long run instead of next quarters profits. If their CPU's are shown to offer significant performance gains over higher MHZ Intel CPU's they might actually gain some ground. I have never used an AMD before but this new chip sounds exciting and I might just wait to upgrade the 'ol pIII until this new chips comes out.

64Bit data movements...

[Anonymous Coward]

For those not interested in the speed increases...

The result of a faster CPU and bus with the ability to move more data is not how we can use programs we have now.

It is about how we can add a new higher level to programming that will run as fast as what we are doing now. Imagine a new language that is as easy to use as Visual Basic but as fast as C++.

Imagine instead your grandmother and grandfather writing 3D games without advanced optimzing tricks and the code runs as fast as Quake 1 does in C++ with an Assembly software render.

Re:64Bit data movements...

[Puu]

So... no benefit to the end user?

Re:64Bit data movements...

[Anonymous Coward]

Just a quick question:

Do you believe in this yourself?

Re:64Bit data movements...

[TENTH SHOW JAM]

Hmm. And the next gen OSs will slow it back down again to where we are today with cruft and bloat.

Just for fun, I installed windows3.11 on a P3 500 and had a boot up in 7 seconds from bios handover to loading up word. I suppose the bit that depresses me is that XP does very little more than 3.1 (more peripherals supported, but not too many more)

And Linux fans, get off your high horse, Mandrake and Redhat on modern hardware cannot compete with aincient distros of slackware for speed of bootage. Yes I know I can roll my own and an out of the box may not be considered "fair" but how many users in the real world do more optimization than is presented during install time?

So more grunt will be sacrificed once again at the alter of bloat.

Sand.

Re:64Bit data movements...

[Anonymous Coward]

Not much more? How ignorant can you get? You get the stupidest thing said on slashdot today award. Congrats!

Re:64Bit data movements...

[Anonymous Coward]

"XP does very little more than 3.1"
Why did you bother to upgrade?

Re:64Bit data movements...

[colganc]

what does booting have to do with the amount of processing power that you have?

why do you even need to turn your computer off? Linux can run for months with out being rebooted...sometimes years. XP isn't as stable, however rebooting it every few days and worrying about losing precious seconds on bootup is crazy.

one more question: how well does windows 3.x recover from errors? when i used it "back in the day" it didn't. was your experience different?

I like you

[Anonymous Coward]

are you friends with YOU FAIL IT?

Re:I like you

[Anonymous Coward]

Click on the secret admirer/stalker link of YOU DID IT! and YOU FAIL IT. That should provide you with the necessary information.

Re:I like you

[Anonymous Coward]

http://slashdot.org/~YOU DID IT!/fans

gotta luv it

[djupedal]

...and anything is can do for people's humor quotient will be welcome as well

i thought...

[subgeek]

i thought that AMD was not going to compete with Intel anymore. so what is the deal with this AMD 64-bit stuff trying to get to the desktop or compete with itanium? or was someone mistaken?

Re:i thought...

[Forkenhoppen]

The article you point to was all about how AMD is interested in branching out. It's no longer their key business, is what they're saying; I wouldn't surprised if they're looking at going up against Motorola, which tends to have it's chips in pretty much everything.

In other words, that article was mostly an appeal for new corporate partners.

As usual, the Slashdot employee got it wrong; timothy, in this case. Funny.. I don't remember seeing him screw up before.

Anyone want to make a scorecard site of some sort for these kinds of things? I think it'd be kinda amusing to keep track of how many times each of the respective Slashdot editors over-exaggerate an article's content. Or, at the very least, it might embarrass them into enforcing some stricter quality control.. (although if last week's two dupe-posts couldn't do it.. ... or was it three?..)

Mod me down if you must, but I did answer the question. :P

64 bits sounds nice, but misses the point

[Anonymous Coward]

64 bits is useful for databases accessing gigantic datafiles and other I/O intensive operations. High performance computing loves moving 64 bit values around. Ever do a "file" on /sbin, /usr/bin, /usr/sbin on Solaris? 32 bits all the way. The only reason I downloaded a trial copy of Sun's C compiler was to compile lsof for Solaris 9, which, since it talks to file structures, needs to be 64 bit.

I think having 64-bit Linux without buying a SPARC, RS6000 or PA-RISC box will be huge for the enterprise. The rest of us will wonder why our apps still suck.

Re:64 bits sounds nice, but misses the point

[Anonymous Coward]

Yeah Sun tends to have timid installers. Have a look in your file system and you will see that 64bit versions of many common utils are available, it just that the 32bit ones are used by default (Yes, even if you did a 64 bit install (This just installs them)).

Anyway back to the 64-bit useful for large databases, etc.

When I bought my current computer 3 years ago (P3-450) it came with 128Mb, while some computers were being sold with 64Mb still.

I'm currently saving for a newer computer (for games) and am looking at the 512Mb range for memory. I have friends that have 700MB+.
Its only a matter of time before we breach 4Gb of standard memory, and this time we will have a nice well worn solution in the form of x86-64 or IA64.

This is much better than back when 640Kb becomes a limit, and Intel didn't yet have a solution.

Microsoft Quote, and Kernel Dev Question

[Erioll]

My favorite part so far (not at the end yet) is the part about why Microsoft would support it: "The relationship [with Microsoft] is very strong, and I think one of the things Microsoft would like to ensure is the tremendous opportunity Hammer presents for them to continue the evolution of the X86 investment is awesome. If you can imagine, put yourself in Microsoft's shoes, you see instead of having to do a right turn, 90 degrees, with the Itanium stuff, now they can see a way of evolving all this investment and technology and continue to be in that dominant position they've had and probably even make it stronger. So Microsoft is obviously selfishly, incredibly interested in making sure Hammer is a success."

The second last sentence is the best. But I like how in the previous paragraph that it says that Linux users are way ahead. Good for us.

As a side-note, any kernel devs out there have anything to say about this new processor? How much benefit do we get out of a 64-bit kernel? How hard is the kernel to port to native 64-bit, and dealing with running old 32-bit apps? I haven't done ANY kernel development myself, but am interested to know what the issues are.

Erioll

Re:Microsoft Quote, and Kernel Dev Question

[nick-less]

As a side-note, any kernel devs out there have anything to say about this new processor? How much benefit do we get out of a 64-bit kernel? How hard is the kernel to port to native 64-bit, and dealing with running old 32-bit apps? I haven't done ANY kernel development myself, but am interested to know what the issues are.

64Bit kernels are available since 1.2 times..

>file vmlinux
vmlinux: ELF 64-bit LSB executable, Alpha (unofficial), version 1, statically linked, not stripped

Re:Microsoft Quote, and Kernel Dev Question

[Erioll]

I know that. But how many of those kernels also had to handle old 32-bit applications as well? I know there have been 64-bit kernels for different architectures, but I'm wondering if this new AMD "feature" causes problems, or if it's trivial to deal with, and is only a re-compile. It's a valid question.

Anybody that has actually coded for it have an answer to this?

Erioll

Re:Microsoft Quote, and Kernel Dev Question

[fastpathguru]

Kernel 2.4.20 has x86-64 support built-in.

Look for SuSE's Andi Kleen in the release-notes.

fpg

Re:Microsoft Quote, and Kernel Dev Question

[Bert64]

The sparc64 linux kernel handles 64bit sparc and 32bit sparc applications, the sparc64 version of solaris does the same. IRIX on 64bit MIPS does too, but i`m not sure about the mips64 port of linux.
Infact... I`m typing this right now on an IRIX64 machine running a 32bit version of galeon:
root@evil:~# file `which galeon-bin` /usr/freeware/bin/galeon-bin: ELF N32 MSB mips-3 dynamic executable MIPS - version 1
root@evil:~# uname -a
IRIX64 evil 6.5 07121149 IP30

Re:Microsoft Quote, and Kernel Dev Question

[Decimal]

My favorite part so far (not at the end yet) is the part about why Microsoft would support it

Back-scratching. Remember how an AMD figurehead testified in favor of Microsoft at the DOJ trial? It may have been enough to convince MS not to walk hand-in-hand with Intel while it tries to crush the #2 chipmaker.

grandpa's geri-porn

[Anonymous Coward]

link, please!

when giblets fly!

[mackstann]

wathing giblets fly all over....is that a reference to eating, or a reference to porn?

640KB should be enough for anyone

[nomadicGeek]

At first they will be expensive, then they will be in the $599 desktops. Why wouldn't you use them?

Re:640KB should be enough for anyone

[Anonymous Coward]

At first they will be expensive, then they will be in the $599 desktops. Why wouldn't you use them?


ummm...

Power Consumption.

Noise from the twin turboprop fans needed to keep the MB from bursting into flames.

The fact that my P90 is still doing fine.

Re:640KB should be enough for anyone

[Anonymous Coward]

The fact that my P90 is still doing fine.

Same here, just looking for an excuse to upgrade...
This may be the excuse.

Re:640KB should be enough for anyone

[swordboy]

At first they will be expensive, then they will be in the $599 desktops. Why wouldn't you use them?

Because 32-bit processors will have evolved into low-power, high-performance, integrated solutions being sold in $199 desktops. 64-bits will flop on the desktop for the same reason the Pentium4 has had so much success - clock speed. Certainly - there will be a 32-bit x86 core on there for backward compatibility, but the extra cruft of the 64-bits will cause much lower clock speeds so, sales will be lower until there are more native 64-bit [desktop] apps. There will be less support for 64-bit apps because sales are low.

In the meantime, the x86 32-bit core will make huge leaps in speed and integration. Low power, fanless desktops are the future. Wireless network booting and the other nine yards...

Re:640KB should be enough for anyone

[jsse]

At first they will be expensive, then they will be in the $599 desktops. Why wouldn't you use them?

I don't know. When I asked my gf why didn't she buy AMD for less money, she said she could pay more for extra stability. Then I said AMD run her XP with a glitch, then she said she wouldn't buy from those sub-brand like AMD. Then I asked why didn't she buy IBM desktop, which is traditionally the origianl brand of PC, then she said it's more expensive...

Things go downhill from here...

Will This be Linux's first killer app?

[peripatetic_bum]

Just wondering, if Linux already runs on 64bit, which I think it does, and I have not heard of microsoft having anything ready for this market, does this mean that just as gamer's buying games pushed the video card (and in my opinion, the os) market, will we see linux be increasing adopted since it will run 64bit and MS does not?
Just a question.
Thanks for the replies

Re:Will This be Linux's first killer app?

[JKR]

...and I have not heard of microsoft having anything ready for this market

MS have been quietly getting ready for 64 bit for at least 2 years; they've been shipping a 64 bit SDK on my MSDN disks for over a year. There are 64 bit NVidia drivers for WinXP-64. What makes you think MS isn't already there?

Re:Will This be Linux's first killer app?

[Anonymous Coward]

I think that Windows 2000 Advanced Server and some version of Windows XP are 64 bit enabled.

Even so, most of the people who will buy 64-bit chips will have been using Linux or will have heard about it.

I think most general desktop users will not switch to a 64-bit CPU, so I don't think that this will be as much of a killer app for Linux until Linux gains mainstream popularity and until 64-bit chips become a mainstream standard.

Re:Will This be Linux's first killer app?

[joshki]

Windows XP and Office XP run on the opteron. There was also an article on slashdot a while back about it. Linux ports are definitely in progress, but I don't know if there's anything solid running on 64 bit x86 machines right now.

Re:Will This be Linux's first killer app?

[mixmasta]

Doh, and let's not forget NT 3.5 ran on Mips and Alpha like ~8 years ago.

Re:Will This be Linux's first killer app?

[Apparition-X]

MS may not have a commercially shipping product, but they certainly have had working code for a long time. I was on a vendor tour 2.5 years ago, and at that time a manufacturer/parnter of MS was already running a 64-bit MS OS on Intel...

Re:Will This be Linux's first killer app?

[guile*fr]

mips wasnt 64bit yet and i think it ran under an emulator on alphas... so it wasnt really 64bits

Re:Will This be Linux's first killer app?

[pantherace]

32-bit version (essentially the top 32-bits got chopped off)

It wasn't bad, A HELL of a lot more stable than ANY x86 windows I have worked with. (As in about as stable as linux on alpha)

Re:Will This be Linux's first killer app?

[Anonymous Coward]

Thanks for the replies

can you say YHBT

Re:Will This be Linux's first killer app?

[mixmasta]

right, .... but it ran on 64-bit hardware. That was me point. =)

Re:Will This be Linux's first killer app?

[fobbman]

Sorry, Microsoft is there, too.

Re:Will This be Linux's first killer app?

[Nothinman]

But the point is moot if it's running in a 32-bit mode. Alpha's are 99% 'normal' hardware with PCI slots, EISA (some with ISA), SDRAM, IDE, etc. The last parts I bought for my Alpha PWS600au was memory from crucial and a PCI video card.

Re:Will This be Linux's first killer app?

[zozzi]

Linux has already been ported to IA-64. A very able David Mosberger and all the other developers have made ported it in time for the IA-64 release (before MS, but no big hurray). There's also a simulator here . Works well enough.

Re:Will This be Linux's first killer app?

[Anonymous Coward]

And windows 3.11 ran on 32bit hardware (386 and up). Also by using only 16bit mode.

64 bit AGP what huh?

[raehl]

You can already do 64 bit adressing on AGP. As for bus width, width of the bus is independent of processor register size. PCI/X already has a 64 bit width configuration and works quite well with processors that are only 32 bits wide.

Wow, the MPAA is *SO* screwed.

[raehl]

These plots are 32 times as good as the 2-bit plots in Hollywood movies.

Good thing it's backwards compatible or all the studios would have to upgrade their writers too.

Re:Wow, the MPAA is *SO* screwed.

[Lozzer]

These plots are 32 times as good as the 2-bit plots in Hollywood movies.

<pedantic>I hope you mean 2^62 times as good if you are talking about 64 bit plots.</pedantic>

The article

[Freston Youseff]

talks virtually nothing about similarities to IA-64. Perhaps, what I'm asking is, can anybody compare and contrast the two architectures; is there a certain advantage to one or the other?

Re:The article

[puppetman]

Perhaps, what I'm asking is, can anybody compare and contrast the two architectures; is there a certain advantage to one or the other?

Yah - AMD will offer it to the consumer combined with motherboards from tier-1 manufacturers like Asus, Abit, IWill, Tyan, and so forth, all at an attractive price (read: the same price as the Athalon XP CPUs).

Intel, on the other hand, will keep their 64 bit CPUs out of the consumer hands by pricing them above what most consumers are willing to pay, thus reaping a premium on them by selling them in servers through Dell and IBM (making even more money on cases and motherboards). There will be limited support for the CPU outside Intel's own motherboard offerings, and if you run with a hard-drive, video card, CD-Rom that has not been explicitly approved by Intel, then forget support (we've had this problem with Intel on some of their server motherboards).

Intel is taking the Cathedral approach, and AMD a Bazaar approach.

Re:The article

[Anonymous Coward]

AMD's x86-64 is an evolutionary approach to 64-bits. It basically is the classic IA-32 instruction set extended to have 64-bit registers, and twice the number of GPRs and SSE2 registers. It is still the CISC on the outside.. RISC core approach than AMD has used since the K5, and internally looks extremely similar to the Athlon.

Intel's IA-64 is a complete break with IA-32. It is a VLIW (EPIC as Intel marketting calls it) processor. Basically, the idea is to simplify the CPU by removing the scheduling algorithms out of the CPU and placing the onus on the compiler to extract the parallelism. The theory is that you can use the transistors you save by removing the OoO Execution, can be used for things such as larger cache or more functional units.

The problem with IA-64 and VLIW designs in general is that it relies heavily on the compiler to extract parallelism, and frankly.. compilers ain't that good YET. The other problem IA-64 has is that Intel has glued on IA-32 compatibility.. which wastes die space that could go to something else, since hte compatibility performs atrociously.. In the original Itanium the IA-32 performance was about the performance of a Pentium 90.

Re:The article

[jpmorgan]

I'd say Intel's strategy of 'consumers don't need 64-bit platforms yet' makes more sense. Why try to sell people an expensive chip that they don't need? They're just going to focus on the very-high-end for a few years.

Of course, that's not to say people won't need to switch eventually... and if you look at Intel's roadmap, they're planning to start weaning people off of x86 and trying to get them to shift to IA64 on the desktop in a couple of years, which considering the current spate of applications and hardware, makes a lot more sense.

Remember, the only real need for 64-bit is larger address spaces, and half of /. was bitching a 18 months ago when WindowsXP came out that it needed at least 128mb of RAM. So it seems to me that we're still a few years away from needing the extra bits of address space.

And until then, Intel will bide their time selling the chip to people who actually have a use for it.

Re:The article

[puppetman]

We have a use for it in our company. As software developers, RAM and GHz is at a premium. We also use workstations as low-end databases for development and QA. Putting 3 Oracle databases on one machine means you needs lots of RAM.

With games, Linux, Oracle, IBM and others porting to match the AMD 64-bit CPU, I think anyone who uses there computer for anything other than email, word processing, browsing the web, will get a nice boost from 64 bits.

Windows runs in 64 bit

[KPU]

Check the Windows XP 64 bit edition website. I hate to burst your bubble, but microsoft knows what it's doing.

Re:Windows runs in 64 bit

[HuguesT]

Microsoft is still not shipping the server versions of XP either on ia64 or x86. We're still at the release candidate stage.

Over 10 years after DEC introduced Alpha ....

[Bio]

I'm amazed to read the discussion, wether or not 64 bit will succeed over 32 bit processors.

This is 10 years after DEC has introduced the Alpha Architecture (in spring 1992).

The Alpha was fun to work with, not only because of it's 64 bit architecture, but because of the clean orthogonal instruction set and it's outstanding performance.

Rest in peace ...

Re:Over 10 years after DEC introduced Alpha ....

[gordon_schumway]

DEAD PERSON: I'm not dead!
CART MASTER: What?
CUSTOMER: Nothing. Here's your ninepence.
DEAD PERSON: I'm not dead!
CART MASTER: 'Ere. He says he's not dead!
CUSTOMER: Yes, he is.
DEAD PERSON: I'm not!
CART MASTER: He isn't?
CUSTOMER: Well, he will be soon. He's very ill.
DEAD PERSON: I'm getting better!
CUSTOMER: No, you're not. You'll be stone dead in a moment.
CART MASTER: Oh, I can't take him like that. It's against regulations.
DEAD PERSON: I don't want to go on the cart!
CUSTOMER: Oh, don't be such a baby.
CART MASTER: I can't take him.
DEAD PERSON: I feel fine!
CUSTOMER: Well, do us a favour.
CART MASTER: I can't.
CUSTOMER: Well, can you hang around a couple of minutes? He won't be long.
CART MASTER: No, I've got to go to the Robinsons'. They've lost nine today.
CUSTOMER: Well, when's your next round?
CART MASTER: Thursday.
DEAD PERSON: I think I'll go for a walk.
CUSTOMER: You're not fooling anyone, you know. Look. Isn't there something you can do?
DEAD PERSON: [singing] I feel happy. I feel happy. [whop]
CUSTOMER: Ah, thanks very much.
CART MASTER: Not at all. See you on Thursday.
CUSTOMER: Right. All right.

Re:Over 10 years after DEC introduced Alpha ....

[Door-opening Fascist]

One thing to remember is that the Alpha was not the first 64-bit processor. Before it were HP's PA-RISC in 1986, U. of Tokyo's TRON design in 1987, and DEC's MIPS R4000 in 1991. Sun/Fujitsu moved the SPARC to 64 bits in late 1992, and IBM was late when it moved the POWER in 1995. So 64-bit processors were neither unheard-of nor new in 1992.

Re:Over 10 years after DEC introduced Alpha ....

[Anonymous Coward]

I'm amazed to read the discussion, wether or not 64 bit will succeed over 32 bit processors.
This is 10 years after DEC has introduced the Alpha Architecture (in spring 1992).

Yes.. and you might have noticed that we're NOT all running Alphas, or even 64 bit CPUs. In fact, we're still running 32 bit CPUs.

So, why is it you're you so amazed to read this discussion? Your two sentences are unrelated, you are not deserving of your +5.
(We're not worthy! We're not worthy!)

Re:Over 10 years after DEC introduced Alpha ....

[msobkow]

IIRC, PA-RISC was 40 or 48 bit addressing with a 32-bit instruction set at the time that Alpha came out. Alpha was out well before 64-bit MIPS processors (which were not DEC processors, BTW.) Alpha was also 64-bit before a refresh of the SPARC design brought it to 64 bit.

Re:Over 10 years after DEC introduced Alpha ....

[Anonymous Coward]

64
1986
1987
4000
1991
1992
1995
???
Duby8 '=w88754onQx87548754o8754mk=8km

Re:Over 10 years after DEC introduced Alpha ....

[Door-opening Fascist]

You're right on both PA-RISC and MIPS. PA-RISC was 48-bit in 1986, and wasn't moved to 64-bit until 1996. MIPS were originally designed by a Stanford research team, but the design was licensed by both Digital and SGI, and a couple embedded companies. I guess I should have done some checking beforehand.

frozencpu.com

[Nazmun]

There are a bunch of nice sites that will show you excellent coolers at under $40 dollars.

Vantec's aeroflow (under 50 dB, i think as low as 38 dB).

32-bit compatible = a temporary half-solution

[justanumber]

No real benefit will come until geniune 64-bit apps hit the consumer market. This will be a steep learning curve for most developers who have only ever know 16 or 32-bit programming.

The problems to be hurdled are:

1) Reliance on the fact that size of pointer is equal to size of int.

2) Reliance on a particular byte order in the machine word.

3) Using type long and presuming that it always has the same size as int.

4) Alignment of stack variables.

5) Different alignment rules in structures and classes.

6) Pointer arithmetic.

A lot of engineering (and developer re-education) work also needs to be put into not only these issues, but also designing the application so that it is actually getting the most out of each clock cycle.

Re:32-bit compatible = a temporary half-solution

[Anonymous Coward]

Of course, those are only C++ problems. Real men use ASM.

But seriously, what would the ASM problems be? As far as I can see, there wouldn't really be any since ASM is nice and low level. No pointers except for the pointer regsiters. No ints or chars, just big data areas.

ASM is so much easier to use than C++. I don't understand why people don't use it. As long as you keep track of everything, it is no more likely to kill your system than C++, and it is so much more elegant when used properly.

Re:32-bit compatible = a temporary half-solution

[Eric Smith]

No real benefit will come until geniune 64-bit apps hit the consumer market.

Given that the Hammer will offer more 32-bit performance per dollar than the current Athlon parts (and, for that matter, Intel parts), I would say that there's at least some benefit even if there are NO 64-bit apps.

But the advantage of Hammer is that you don't need to migrate ALL of your apps to 64-bit to get a serious performance benefit. With the IA64, the performace of 32-bit applications is terrible, so it's a poor choice unless most of your software is 64-bit.

Re:32-bit compatible = a temporary half-solution

[MagPulse]

Developer re-education? Developers constantly re-educate themselves or become outdated quickly. Most C/C++ programmers who will be affected are aware that if we don't make our programs 64-bit clean today that we'll pay for it later.

C99 has inttypes.h, which declares types like int32_t and uint16_t. Bjarne says this will probably make it in to C++0x.

Re:32-bit compatible = a temporary half-solution

[justanumber]

If I buy any architecture in the future, by Moore's Law and the scale economies of the chip market, it's going offer better price/performance. This would be true for a 32-bit architecture or 64-bit or n-bit.

So what's your point?

Until consumer development shops start refactoring and releasing genuine 64-bit apps rather than just porting them (as I described), the benefits of 64-bit for the consumer will be negligble.

Re:32-bit compatible = a temporary half-solution

[justanumber]

Agreed. but it's more than making your code 64-bit clean.

A naive port simply carries source code over, re-compiles and re-links it as necessary, and uses the executables that are verbatim translations of their 32-bit counterparts. This should, in general, give correct functionality.

Unless you optimize to take full advantage of features that the new architecture has available like parallel instructions and introspection, your performance can actually be worse.

Re:32-bit compatible = a temporary half-solution

[Eric Smith]

The benefit will be higher performance, whether the user runs 64-bit apps or not, for the reasons I stated. It's a simple fact that the Hammer will offer higher performance on 32-bit applications than the 32-bit Athlon, because it's a newer design. AMD can't get as much performance improvement from simple process geometry scaling on the 32-bit part as they can from the redesigned core of the Hammer. This has nothing to do with the 64-bit-ness of the Hammer. If Moore's law was enough by itself, Intel wouldn't have had to make the Pentium IV; they would have just kept scaling the Pentium III (or II, or I, or the 486...). So that doesn't seem like a negligible benefit to me.

Also, "porting" and "refactoring" aren't necessary. Just recompile. All your 64-bit operations (e.g., operations on ISO C int64_t and uint64_t types) get faster. And you can access more per-process memory. That's really the point. There's nothing magic about 64-bit mode that requires you do do amazingly special stuff to take advantage of it.

If you were somehow expecting using 64-bit pointers to make your software (that fits OK in under 4G address space) run faster, think again.

If you're proposing some other sort of speedup from native 64-bit applications, you'll have to explain what you think is going to provide that.

Re:32-bit compatible = a temporary half-solution

[cartman]

1) Reliance on the fact that size of pointer is equal to size of int.
2) Reliance on a particular byte order in the machine word.
3) Using type long and presuming that it always has the same size as int.
4) Alignment of stack variables.
5) Different alignment rules in structures and classes.
6) Pointer arithmetic.

With regard to points 1, 2, and 3: Only exceptionally hacky programmers would ever do these things. Even the first edition of "The C Programming Language" states clearly that you should never do them. It also states that you should always use "sizeof (int)" rather than 4 (32 bits), etc. It also states that a long should never be assumed to be the same size as an int. Only very poorly written programs would present any difficulty when re-compiling for 64 bits.

Points 4, 5, and 6: These things are handled automatically by the compiler. If an int is 8 bytes, the compiler automatically multiplies by 8 rather than 4.

Re:32-bit compatible = a temporary half-solution

[Ninja Programmer]

Yes, but several development houses (including, primarily Microsoft) have had experience porting to other 64 bit architectures. So the problems are known and addressable.

• The problems to be hurdled are:
  1) Reliance on the fact that size of pointer is equal to size of int.

True. But this is really a bad code practice sort of thing that developers that the C/C++ standard doesn't endorse anyhow. Java developers need not be concerned.

• 2) Reliance on a particular byte order in the machine word.

Its little endian, like it has always been on Intel. What's the problem? Little endian has the natural LSB/W/D property that makes increasing the natural word size typically a non-issue (unlike big endian where it is a serious PITA.)

• 3) Using type long and presuming that it always has the same size as int.

It is the same. You need to use "long long", or "_int64" or something like that to move to 64 bits in your C/C++ compiler. AMD's new "long mode" actually defaults to 32 bit data sizes, and only uses 64 bits when specifically overridden to do so. That's the whole point of the x86 architecture -- it supports a variety of data sizes as a consequence of its long history of backward compatibility, not just one (like a typical RISC.)

• 4) Alignment of stack variables.
  5) Different alignment rules in structures and classes.

Same as it has always been. (64 bit integers already exist today in known common ABIs/compilers, in case you were unaware.)

• 6) Pointer arithmetic.

Eh? You can add/subtract integers to pointers, and subtract two pointers from each other. How does moving to 64 bits change anything?

Re:32-bit compatible = a temporary half-solution

[WatertonMan]

Many, although not all, of those problems can be solved by compiling your key code on multiple platforms. Obviously you'll have some system dependencies - especially in the interface. But it is almost always a good idea to isolate the interface from the "substance." Making the "substance" crossplatform is almost always a good idea, and not just because of future sales or development.

If you do this sort of crossplatform development you'll find bugs easier and most of the above sorts of issues never happen. Then when 64-bit computing arises, you'll have few problems in your development. I know I compile under OSX, Linux, Solaris, Borland and Visual Studio. It really does find a lot of bugs and prevents a lot of problem. I do a cross-compile every few weeks and it is amazing how many things pop up that never did in my main development system.

Re:32-bit compatible = a temporary half-solution

[be-fan]

It'll benifet Linux users like me right off the bat. Most Linux software is already 64-bit clean. The Alpha port of Gentoo was done in *two* days! I'm a recompile away from enjoying the fastest processors out there, regardless of architecture. Suck it Gates!

Re:32-bit compatible = a temporary half-solution

[rweir]

Aren't these all good practice tips anyhow? There's lots of crappy developers producing crappy 32-bit-specific code out there, but it's because they're crappy developers.

Re:32-bit compatible = a temporary half-solution

[timeOday]

I was on a C++ project on Solaris that switched from 32 bit to 64 bit half way through. It was a piece of cake. In fact you could just pass a compiler switch to compile for one or the other.

Re:32-bit compatible = a temporary half-solution

[cerberusti]

Actually there are valid reasons to rely on the byte order in the machine word. Alignment of stack veriables also has valid reasons (though not as frequently). The fact it is relied upon in cases it should not be does not affect the fact that there are some instances for which it is a good idea to. For an example, a library I wrote recently (dealing with graphics) relies upon the byte ordering quite heavily, although a define controls which byte ordering is assumed. Without assuming byte ordering, the speed of the library would decrease by about 50%, which is unacceptable as it is designed for use in real time. This falls into the same category as goto statements, most of the time they are used in situations where they should not be but, there are times when using them produces cleaner code and speeds up the program. Keep in mind that C is a fairly low level language and, there are cases in which you can use it instead of asm to gain some portablility and for some of those uses, some of these things really matter. Exceptionally hacky programmers may do these things frequently but, some very good programmers do some of them occasionally for good reason.

Re:32-bit compatible = a temporary half-solution

[cerberusti]

C++ and C especially are far more portable than ASM. If you are targeting multiple architectures, ASM would require multiple versions. Newer programmers also usually are not very good even when the claim to know ASM (you would not believe the number of times I have had to yell at people about assigning zero to a register), and usually people who are decent at ASM are far more expensive than their counterparts doing C, C does a fairly good job of optimizing despite the skill level of the programmer. That being said, yes, real men do use ASM. (-:

Re:32-bit compatible = a temporary half-solution

[Anonymous Coward]

try writing a loop that requires 8 registers to run nicely while you have only 7 around..
it hurts less in hll when you can't see the wasted cycles.

Re:32-bit compatible = a temporary half-solution

[Anonymous Coward]

Eh? You can add/subtract integers to pointers, and subtract two pointers from each other. How does moving to 64 bits change anything?

char *p1=(char*)0x123456789abcd;
char *p2=(char*)0x3456789abcdef;
int d = p2-p1;

ok. the result is supposed to be quite undefined too but you never know what kind of hacks you run into..

Re:32-bit compatible = a temporary half-solution

[rnd()]

these are all good arguments in favor of languages like Java and LISP (and maybe eventually C#).

Non embedded-systems programmers should not waste time worrying about machine level details, or even worrying about using best-practices to ensure cross-platform code, unless performance is more important than maintainability and reliability.

I have only one thing to say...

[El Camino SS]

I have only one thing to say to all of you people that say that we all have too much home processor power:

SimCity 4.

Re:I have only one thing to say...

[AbRASiON]

I must say the screenshots were *REALLY* detailed... I wouldn't be surprised if this game required 4gb of ram to access all the little sprites / sounds / textures / details in ram

*(ok ok I lie, but it's a 256->512mb game, more so 512 I'd say from the shots)*

Damn do I want a killer app for 64bit, something I can't *DO* now- I want a REASON to buy one and I want something COOL I couldn't do before - they make that and 64bit = super success.

(example if Doom 3 was 64bit only) ...

AMD is puny

[Ed Avis]

From the interview:

We really can't control whether we'll go to war with Iraq, and all that sort of thing.

And that, my friends, is the difference between AMD and Intel.

Re:AMD is puny

[Puu]

Jesus. This Hector Ruiz guy makes Ellen Fleiss look sober.

Re:AMD is puny

[Anonymous Coward]

I think what he's trying to say is that the performance of the economy (and with it demand for AMD cpu's) will depend on whether or not we go to war with Iraq, because that prospect will have a big influence on oil prices. Idealogy isn't the issue. He's a businessman whose success is determined by his ability to adjust to the economic climate.

Too late

[Anonymous Coward]

>>>
but there really is no need for the general public to have 64 bit processors.
>>>

Last time I checked, digital video output take up more than 4GB (the 32-bit limit). The key thing is 64-bit microprocessor 64-bit addressability which makes certain class of applications practical. Spreadsheet and wordprocessors were 32-bit class applications even when we were using 16-bit microprocessors. We had to fake addressability using segment pointers and expanded memory. Yuck! With digital output bigger than 4GB, it is not trivial to randomly access different portion of the file like accessing a pointer using 32-bit addressing.

Yes, we know that accessing a datum accross 64-bit space will take forever even if we have 2000GHZ microprocessor. Just because we can access 64-bit address space does not mean we have to use/populate all of them. With sparse addressing we can use faster memory allocation method that does not have to worry about compactness and fragmentation.

Later

"comparatively priced"?

[SkulkCU]

AMD's 64-bit chips will be comparatively priced to the 32-bit ones

So, they're going to be twice as much?

heh.

Re:"comparatively priced"?

[Anonymous Coward]

So, they're going to be twice as much?

Do you mean to ask if these 64 bit chips will cost about the same price as the current 32 bit Intel offerings?

Re:"comparatively priced"?

[spike hay]

AMD's 64-bit chips will be comparatively priced to the 32-bit ones

So, they're going to be twice as much?

No, they'll cost 2^32 times as much.

Well... there was Alpha

[chunkwhite86]

The kernel is already 64 bit compatible. Remember Alpha? I have a PersonalWorkstation 500Mhz Alpha. It's running RedHat 7.2 and it's great.

Thats one of the great things about our linux kernel and many GNU applications - They've been running in 64 bit land on the Alpha for years.

While others will have to do ports and worry about the little-endian and big-endian details, We're already there as far as the software is concerned.

Re:Well... there was Alpha

[Anonymous Coward]

Don't forget that RC3 of Windows NT 5 (later called Windows 2000) included a release for the Alpha chip.

Now that Intel has been given the Alpha chip by Compaq, the dev work for Microsoft is fairly small for porting to the Itanium 2. IIRC, the Itanium 2 is the same as the EV9 Alpha Processor, so most of the code written for the earlier Alphas could be ported to the newer chipset w/o much of a problem.

Re:Well... there was Alpha

[Anonymous Coward]

As has been pointed out. NT 3.5 and 4.0 were also run on the Alpha and other 64 bit chips. Windows the OS and windows software have been running 64 bit for over a decade!

Re:Well... there was Alpha

[chez69]

windows NT ran in 32 bit mode.

Re:Well... there was Alpha

[Bert64]

If only this were the case, the itanium 2 is very different to the alpha.
Really they should have continued the alpha, instead of creating a new architecture... The alpha is the cleanest of all the 64bit architectures, and has always been the most performant, plus by using an existing architecture you would already have a software and user base.

re: Skip to the last page for the most interesting

[bogie]

No this is the interesting stuff

"eWEEK: What does it mean to you personally, though, when a Gateway or an IBM not just stop, but announce that they'll no longer be offering AMD as an option?

Ruiz: I think it's terrible, obviously. It's terrible. I think if you were to talk with Ted Waitt at Gateway, and ask him, "Why'd you do that?" and if he would really tell you why, it's a question of he's being bribed to do it. Now, he's got to look out for his own hide and the company that's probably in great difficulty has got to listen to the huge amounts of money that can help him do that.

But you know what I find amazing, think about the power, is that despite all that, which obviously we really get emotional about the fact that somebody like Gateway gets bribed into doing that, is that despite that, according to Dataquest last week, we're still holding a 19 percent share of the market. That to me tells me we're in the throes of breaking this open"

Hey Intel, see you in court! Of course now that Intel is along with Microsoft backing a group to outlaw opensource in the government, I think its time for the opensource community to boycott Intel. Why should our money go to a company which is now attempting to hurt Linux and opensource? I know because these recent actions, I will NEVER buy Intel ever again!

So in other words...

[hendridm]

> He says that AMD's 64-bit chips will be comparatively priced to the 32-bit ones, and backwards compatible.

So in other words, the same chip with a new label? Or just the same perfomance?

64 bit

[Konster]

64bit CPU, 64bit color depth in my Radeon 12500/64 Pro, 64bit bus! Doom64 with 64bit giblets!

What did Bill Gates say about bits a few years ago? That 64bit ought to be too much for everybody?

Re:64 bit

[guile*fr]

u dont want to play doom64 it sucked...

Remember the Alpha

[chunkwhite86]

I see many posts here wondering about porting Linux to 64 bit...

Remember the Alpha? 64 bit goodness all the way. Has been running Linux for years.

And for those old enough to remember... Microsoft did support Win NT on the Alpha just a few years ago.

As far as the software goes, both Linux and Microsoft are ready for 64 bit computing.

Re:Remember the Alpha

[puppetman]

Yah, but the alpha was technology out of the hands of the normal user. And you can't buy a kick-ass 3D video card for it.

And I bet the ogg/mp3 tools and players don't exist or suck.

Linux-64, and Unreal Tournament 2003 re-compiled, plus support for those 32-bit games and apps means there will be an instant software base.

Re:Remember the Alpha

[Anonymous Coward]

Why would you want a recompiled version of ut 2003, compiling a program designed for 32-bits, on 64 will just make the program run slower.

Re:Remember the Alpha

[Bert64]

Alpha`s used to ship with PowerStorm displaycards a few years ago, which at the time were able to completely destroy any pc displaycards in any benchmarks..
Modern alpha`s support AGP and PCI.. you could plug any displaycard in assuming you have drivers for it, and most drivers written for linux/x86 work fine on linux/alpha

Re:Remember the Alpha

[puppetman]

Huh?

Epic said:
"We're extremely excited about the upcoming AMD Athlon 64 processor. It's designed to run today's 32-bit games and applications and will actually, clock-for-clock, boost their performance relative to 32-bit AMD Athlon processors. But when you add 64-bits to the picture you can get additional capabilities and other performance boosts," said Mark Rein, vice president, Epic Games Inc. "We're planning to ship a 64-bit update for Unreal Tournament 2003 at the same time the AMD Athlon 64 processors show up on retail shelves."

Anandtech said:
"Epic has said that they will ship a 64-bit version of Unreal Tournament 2003 when Athlon 64 makes it to retail, and the performance increase from moving to Opteron is significant. You get a 15% speedup from the on-die memory controller and another 15% from compiler optimization for the additional registers in x86-64 mode."

Just to remind people why more bits is good..

[Inoshiro]

2^32 addressing is obsolete already -- it cannot keep up. Most enthusiasts have a gig of RAM (or more) in their DESKTOP PCs. In 2005, most of them will have hit the 4gb limit. In 2009, most consumer PCs will have hit the same limit. Servers have already hit this limit. That's why there are special instructions (a return to segmented memory access) on P3 and P4 processors, allowing up to 64gb of RAM in 4gb segments to be addressed. If you remember doing DOS programming (I do), you know why this 64-bits is good, while 32-bit segmented access isn't.

2^32 addressing limits addressable HD space to 2 terabytes. "2 terabytes? But that's way larger than even enthusiasts use in their PCs, despite their larger than average needs." This ignores the fact that many companies have storage arrays that are at 2 terabytes. Some work went into the 2.5 Linux kernel to increase the number of blocks that could be addressed by moving internally to 64-bits. Storage needs are always increasing. If we're hitting 2tb today, isn't it a good thing that we're moving to a better amount of bits?

2^64 addressing is not the only benefit of the change. FPUs see additional benefit when they have more bits. More bits means more precission; this is very important and desirable, especially when working with numbers that have fractional components. For proper 3D rendering, physics models, and anything else that involves computing numbers that have fractional parts, more is better. When the FPU can handle a double in one clock cycle because it works natively on 64-bit IEEE floating point numbers, you will notice a performance boost in addition to the increased accuracy.

64-bit word operations means that databuses can be slower, since each clock-tick sends more data. 64-bits means you can do more, more flexibly, with your computer.

There will always people who resist change, even when there is no reason to resist change. The same people are posting comments on Slashdot about how 32-bits is enough, and how happy they are with 32-bit applications. These are the same people who had to be carried, kicking and screaming, from their 286s to the new 386 and 486 machines which had 32-bit addressing and data operations. Don't let these people hold back your exploration of new technology!

For those of you who are saying, "what about 64 bits? Will 64 bits be enough?" 2^64 is 32 orders of magnitude bigger than 2^32. 2^32 is roughly 4.5 billion (unsigned). 2^64 unsigned is 18,446,744,073,709,551,616, or roughly 2220 * 8309 trillion. 4.5 billion goes into that number 4.5 billion times. 2^64 is certainly enough for at least a hundred years :)

Re:Just to remind people why more bits is good..

[Rob.Mathers]

"2^64 is certainly enough for at least a hundred years"

Famous last words?

Re:Just to remind people why more bits is good..

[Christopher Thomas]

2^64 addressing is not the only benefit of the change. FPUs see additional benefit when they have more bits. More bits means more precission; this is very important and desirable, especially when working with numbers that have fractional components. For proper 3D rendering, physics models, and anything else that involves computing numbers that have fractional parts, more is better. When the FPU can handle a double in one clock cycle because it works natively on 64-bit IEEE floating point numbers, you will notice a performance boost in addition to the increased accuracy.

Um, all current x86s already handle 64-bit IEEE double-precision floats natively (actually more like 80 bits, for "extended double-precision"). The FP register file has been this wide for quite a while.

There will be no performance or precision boost for floating-point math from moving the rest of the chip to 64-bit registers/datapaths.

Re:Just to remind people why more bits is good..

[blamanj]

A nit. Orders of magnitude is generally thought of in the decimal realm. Thus 2^64 which is a 20 digit number is only 10 orders of magnitude greater than 2^32 (a 10 digit number).

I wouldn't be to sure about the 100 years part either. But it out to be good for at least 10.

Re:Just to remind people why more bits is good..

[Anonymous Coward]

Moore says 32 extra bits will last about 48 years.

Re:Just to remind people why more bits is good..

[cheezedawg]

2^64 addressing is not the only benefit of the change

I don't know of any of the 64 bit procs that actually use 64 bit addresses yet. AMD's Hammer will only use 40 bit addresses IIRC.

Re:Just to remind people why more bits is good..

[Anonymous Coward]

No- the density of transistors on silicon has little to do with this (and that is all Moore has talked about)

Re:Just to remind people why more bits is good..

[Anonymous Coward]

Double the density of transistors and you get twice the memory in the same space. Twice the memory means you need one more bit to address the additional memory. One bit every 18 months. 48 years.

Re:Just to remind people why more bits is good..

[lucifuge31337]

2^64 is certainly enough for at least a hundred years:)

"I think there is a world market for maybe five computers."
- Thomas Watson, chairman of IBM, 1943

"I have travelled the length and breadth of this country and talked with the best people, and I can assure you that data processings is a fad that won't last out the year."
- The editor in charge of business books for Prentice-Hall, 1957

"But what...is it good for?"
- Engineer at the Advanced Computing Systems Division of IBM, 1968, commenting on the microchip

"There is no reason anyone would want a computer in their home."
- Ken Olson, president, chairman and founder of DEC

Come on....you HAD to know that was coming.

Re:Just to remind people why more bits is good..

[Azog]

64-bit really should be enough for a hundred years. If you could turn every atom of the earth into a bit of storage, 64 bit addressing would still be enough to give every byte a unique address. (or something like that.)

So, unless we find a way to make computers without atoms (memory made entirely of energy in some sort of star-trek-esque subspace lattice?), or, it becomes common for single computers to contain as much matter as a planet, 64 bits will be enough...

I'd bet you a petabyte of memory we'll be fine for 100 years :-)

Re:Just to remind people why more bits is good..

[user32.ExitWindowsEx]

With IPv6 we could give eveyr atom its IP address!!!
Just think! CmdrTaco could turn Slashdot into a [Dr. Evil]"Death Star" with a "laser."[/Dr. Evil]

Re:Just to remind people why more bits is good..

[leviramsey]

I wouldn't be to sure about the 100 years part either. But it out to be good for at least 10.

32-bit was good for 10+ years. If memory demand is growing in some Moore's Law-like fashion (which is probably somewhat reasonable), 64-bit should be good for at least 20 years, though when you start the clock matters...

Re:Just to remind people why more bits is good..

[Anonymous Coward]

Bzzt. You're thinking of IPv6 addresses, which are 128 bit wide. 2^64 addresses wouldn't be enough to address the atoms in a few sheets of paper.

Re:Just to remind people why more bits is good..

[Zaak]

The physical interconnect is of secondary importance to the internal implementation. If your program counter and other registers have 64 bits internally, then to make a processor which can actually use 2^64 bytes of memory, you just need to add more address lines to the IC. No big deal. When your registers are only 32 bits (as they are in the IA32 processors we have now) it's not easy to make a processor which can use more than 2^32 bytes of memory. You have to use icky segmentation schemes and other ugliness.

TTFN

Re:Just to remind people why more bits is good..

[Amazing Quantum Man]

The FP register file has been this wide for quite a while.

Since at least 1981 -- the 8087 had 80-bit FPU registers.

Re:Just to remind people why more bits is good..

[kma]

Servers have already hit this limit. That's why there are special instructions (a return to segmented memory access) on P3 and P4 processors, allowing up to 64gb of RAM in 4gb segments to be addressed.

Bzzt. The feature you're describing is known as PAE, for physical address extension. It doesn't work via "real mode" style DOS segmentation. Each program's virtual address space is still 4GB, and pointers are still a flat 32 bits. PAE simply changes the hardware page table structure so the 4GB "window" of your virtual address space can look out onto more than 4GBs of physical memory. Even though no one process can access more memory than before, you can run multiple, 4GB processes on a single machine.

Miraculously, someone at Intel stowed the x86 crackpipe, preventing some sort of segmented/overlay nightmare like the one you describe.

Re:Just to remind people why more bits is good..

[Sloppy]

A few sheets of paper should be enough for anybody.

Re:Just to remind people why more bits is good..

[Luminous Coward]

If you could turn every atom of the earth into a bit of storage, 64-bit addressing would still be enough to give every byte a unique address.

Not quite.

64 bits encode approximately 2*10^19 values.

On the other hand, there are between 10^48 and 10^51 atoms in the earth.

You, my friend, are off by a few orders of magnitude.

Re:Just to remind people why more bits is good..

[Vegan Pagan]

"18,446,744,073,709,551,616 is certainly enough for at least a hundred years :)"

Microsoft is putting that much money into making sure the next version of Windows only lets it last for 100 weeks.

Re:Just to remind people why more bits is good..

[Weirsbaski]

That's why there are special instructions (a return to segmented memory access) on P3 and P4 processors, allowing up to 64gb of RAM in 4gb segments to be addressed.

If you're referring to the 32-bit RETF instruction, it's been around since the 386 days. And it still returns to a 32-bit address, because only 4gb of RAM is visible at any one time- the OS has to explicitly change the page-tables to map other physical memory into the 4gb virtual space.

2^32 addressing limits addressable HD space to 2 terabytes.

Refresh my memory, how does the cpu's 32-bit limit on virtual addresses control the addressable HD space? The OS may have created such a dependency, so fix the OS.

2^64 addressing is not the only benefit of the change. FPUs see additional benefit when they have more bits. More bits means more precission; this is very important and desirable, especially when working with numbers that have fractional components.

But hammer doesn't add more precision bits to the x87 or SSE registers. It does, however, add _more_ sse registers. And, of course, the FPU instructions will be able to reference tables in memory that are >4GB long, if such a need arises.

I think hammer will be the greatest thing since diced cheese, but not for all the reasons above.

Re:Just to remind people why more bits is good..

[Luminous Coward]

Each program's virtual address space is still 4GB, and pointers are still a flat 32 bits. PAE simply changes the hardware page table structure so the 4GB "window" of your virtual address space can look out onto more than 4GBs of physical memory. Even though no one process can access more memory than before, you can run multiple, 4GB processes on a single machine.

As far as I understand, it is possible to run multiple 4 GB processes on a 32-bit machine with virtual memory (although swap would need to be very large). PAE would only allow one to have more than 4 GB of RAM?

Re:Just to remind people why more bits is good..

[RJHill]

If you remember doing DOS programming (I do), you know why this 64-bits is good, while 32-bit segmented access isn't.

Heh. Here's an old joke.

Q: What's the difference between an IBM PC and a boat anchor?
A: Segment registers.

If you know what this means, shouldn't you be off playing shuffleboard or something?

Re:Just to remind people why more bits is good..

[Nothinman]

But it's still a bad hack because:

A) Some programs need >2G of space (with some versions of NT you can boot with the /3G switch to get 3G of addressable space, but virtually no apps support it.
B) Some devices, mostly ISA but some dumb PCI ones, can only DMA to addresses below the 4G line.

Re:Just to remind people why more bits is good..

[firewood]

There will be no performance or precision boost for floating-point math from moving the rest of the chip to 64-bit registers/datapaths.

IEEE doubles only have 52 bits of mantissa, so 64-bit scaled integers actually have more precision (though less dynamic range). 64-bit long longs are also very useful for building quad/oct precision integer or floating point math libraries.

Re:Just to remind people why more bits is good..

[G-funk]

The pentium pro onwards has a 40 bit address bus.

Re:Just to remind people why more bits is good..

[cerberusti]

Miraculously, someone at Intel stowed the x86 crackpipe, preventing some sort of segmented/overlay nightmare like the one you describe.

no, the guys doing IA64 stole it, rest assured, it is still being used.

Re:Just to remind people why more bits is good..

[crgrace]

For those of you who are saying, "what about 64 bits? Will 64 bits be enough?" 2^64 is 32 orders of magnitude bigger than 2^32. 2^32 is roughly 4.5 billion (unsigned). 2^64 unsigned is 18,446,744,073,709,551,616, or roughly 2220 * 8309 trillion. 4.5 billion goes into that number 4.5 billion times.

Wouldn't that be 9.5 orders of magnitude?

Re:Just to remind people why more bits is good..

[Anonymous Coward]

Erm, you don't obviously don't understand virtual memory. First of all, it's partially implimented in hardware, so no shit it's gonna run on virtual memory. Secondly, swap wouldn't need to be very large, swap could be tiny, pages are only paged out to swap when you're running out of RAM, so if you have gigs and gigs of RAM you don't have to mess with much swap, do you?

Re:Just to remind people why more bits is good..

[bored]

Miraculously, someone at Intel stowed the x86 crackpipe, preventing some sort of segmented/overlay nightmare like the one you describe.

No, its just the rebirth of an even more evil concept, "bank switched memory" because that is effectivly what has to happen to use more than 4 gigs at a time. Look at the W2k "Address Window Extensions" which allow more than 4gb access in a manner similar to the old DOS EMS extension. In order to support this they have to be constantly fsking with the page tables, which results in lots of TLB flushes, and on SMP boxes IPI's. In the end its really slow slow slow! Plus you end up supporting the thing as though you have a bunch of 4k segments.

More bits not useful to games?

[Inoshiro]

Have you ever done a physics engine? When you are working with vectors, you want as much precission as you can get. More precission means more bits.

Re:More bits not useful to games?

[Anonymous Coward]

Internet Explorer (n): Another bug -- that is, a feature that can't be turned off -- in Windows.

IE can be uninstalled. Try add/remove windows components. BTW, are you the Inoshiro from Kuro5hin?

Re:More bits not useful to games?

[Anonymous Coward]

That he is, and more! I'm "glad" he was removed as a poll-option; makes the K5 gang look better than CowbodyNeal-still-stuck-as-an-option.

Re:More bits not useful to games?

[MajroMax]

Have you ever done a physics engine? When you are working with vectors, you want as much precission as you can get. More precission means more bits.

Sure, but precision happens in the floating point units (and SSE(2) vector unit); 64-bit integer registers won't help with float precision.

Re:More bits not useful to games?

[Anonymous Coward]

then you're screwed.

do
for (...) y=rotate(x, i*k);
instead of
for (...) x=rotate(x, k);

etc.
floats are a lossy format. treat them as such.

Excellent

[Anonymous Coward]

So does this mean I can buy a motherboard from SiS or VIA and have a 64 bit AMD cpu all running on non-palladium hardware? I'll install Slackware linux or FreeBSD and I'm set!

Re:Excellent

[stratjakt]

AMD is supporting palladium too.

So, no.

Desktop advantages of 64 bit

[chunkwhite86]

A major advantage, especially to the Open Source community of 64 bits on the desktop is software development.

Remember, many (most?) open source developers are private individuals and not huge corporations. Allowing individual open source developers to own an affordable 64 bit desktop machine will allow them to more effectively develop and debug the code that runs on the 64 bit servers.

It only seems natural that a developer, given a 64 bit system to develop and debug code on, is going to produce better 64 bit code. And we all want Linux (and the BSD's!) to be the best 64 bit platform it can be, right??

Alas, the memory...

[NerveGas]

I've been eagerly waiting for Hammer ever since it's announcement. High-bandwidth interconnects, 8-way SMP support, and AMD's incredibly high IPC all team up for a chip that sounds like a winner.

However, each chip is only going to get a single DDR333 memory path. With all of this time and effort, and so much at stake for AMD, you'd think that they'd make sure that they did it right, and move to a dual-channel solution, or at the very least, a DDR400 solution - which will be a pretty standard offering when the Opteron/Hammer/Athlon64/Whatever is released.

Sure, it'll perform pretty well with a single channel of DDR333. But I'll bet it would perform MUCH better with more bandwidth. And compared to all of the design and development that they've already done, implementing a dual-channel memory controller really wouldn't have been any significant challenge.

So, I'm not nearly as optimistic. On the other hand, I'm not a skeptic yet. When they come out, I'll see how they perform. But I'm certainly not as excited as I used to be.

steve

Re:Alas, the memory...

[fastpathguru]

Opteron has two DDR memory interfaces. Happy now? fpg

Re:Alas, the memory...

[Anonymous Coward]

I've been eagerly waiting for Hammer ever since it's announcement. Last 200 years must've been pretty boring, then. Release the fucking CPU within my lifespan, AMD.

Re:Alas, the memory...

[Anonymous Coward]

Maybe, but AMD is probably running late in getting their product out the door. That's bad.

Re:Alas, the memory...

[orz]

The high end Hammer chips will have either a 128-bit DDR333 memory bus or 2 64-bit DDR333 memory busses. Either way, that's 5.4 GB/s of memory bandwidth (per CPU), compared to the 2.4 GB/s that you get with a normal 64-bit DDR266 memory bus. The low end Hammer chips will admittedly have merely 2.7 GB/s memory bandwidth (per CPU), but most desktop applications aren't limitted by memory bandwidth. What are you complaining about?

Re:Alas, the memory...

[be-fan]

Well, on the desktop, memory bandwidth is extremely important. Thanks to SIMD, current P4's can chew through tens of GB of data per second. In an intensive program like a game of media application, even the 3.2 GB/sec of memory bandwidth of a P4 can get pegged. Intel is moving to dual channel DDR in all of its chipsets, and the KT400A will also be a dual channel platform. Starting in Q2-2003, Intel's SpringDale-PE chipset will have 6.4 GB/sec memory bandwidth, and people buying DDR-433 memory will probably be able to push the KT400A to 7 GB/sec of memory bandwidth. All for less than $500 on PriceWatch :)

A marginal OS on a marginal processor?

[Kjella]

What does it get the consumer:
New apps (as in killer apps)? No.
New OS features (by going 64bit)? None.
Speed? Somewhat.

Since when did a little bit more speed make linux a killer app? Also considering that if there is a marked, Windows will most certainly give out a 64bit-version.

Kjella

How do you build one?

[astrashe]

Are there any sites that talk about how to get a 64 bit AMD system going? How expensive are they?

NT on Alpha

[Detritus]

NT ran in a crippled 32-bit mode on the Alpha. It did not support 64-bit applications.

Re:NT on Alpha

[WinterSolstice]

Yes, there were major issues. Calc and notepad were about the only apps that came with it, and nothing would run on it except for a very short list of apps.

Or wait, was the the x86 version of NT 4.0...

Just kidding, of course. We actually had an NT 4.0 on Alpha here for a long time, but we never actually used it for anything. It was a sort of expensive air filter after the original project (whatever it was) dried up.

-WS

Re:NT on Alpha

[humanasset]

There were a fair amount of other applications for Alpha-based NT boxes. If I remember correctly, MS SQL and Exchange both supported the Alpha platform. There also was a wide variety of other applications, from PTC's Pro/ENGINEER to the accounting package Great Plains Dynamics, that also supported the Windows NT Alpha platform. Microsoft even had Alpha native versions of Word and Excel 97.

Even if there was no native Alpha application, you could still install and run most x86 applications using DEC's FX!32 x86 emulator. It ran at about half the speed. But this is when DEC had 1 GHz Alphas -- about five years ago! The fastest processor from Intel was only about half that.

Having a dual-boot Alpha with NT and Digital UNIX was a real treat!

HIT THE DECK, INCOMING 'S'!

[Anonymous Coward]

Who's this "Benchmark" character? What does he have that you're referring to?

What about the OS

[James Littiebrant]

A 64-bit may be great but a 64-bit OS is needed to fully utilize that power. Windows may not be prepared yet for a 64-bit processor and linux will also need an over haul also. Is the world ready to accept a 64-bit processor? for now it is not. Only time will tell if it is.

Re:What about the OS

[cant_get_a_good_nick]

Windows 3.51 ran on Alpha, Mips, and PowerPC. 4.0 was restricted to Alpha only, then some service pack took even this away. Even though it doesn't run on anything but IA32, I doubt if the code is real "64-bit impure" now, and they probably already have had x86-64 and Ittanium teams chugging along for a while now. Most of the free UNIXes (Linux, FreeBSD, NetBSD) have Alpha support, which is another 64-bit Little Endian chip, so it shouldn't be too hard of an shift to x86-64. NetBSD already has the OS up and running on an x86-64 simulator.

If you want to say "an OS optimized for 64 bit" and applications that scream with 64 bit chips, then I'd say "ahh, there's the rub." It will take folks to move to 64 bit, just as it took folks to move from the relatively dirty 16-bit "selector plus offset" addressing of Intel 16 bit code to the cleaner flat addressing model of 32 bits. but you need the chips out there for people to play with and get real applications for.

Re:What about the OS

[Anonymous Coward]

I'm sure we'll run 32-bit cpus for at least a 1000 years to come. no?

Re:What about the OS

[Anonymous Coward]

Nah, I'm hoping for an asynchronous logic-based cpu "thousand year reich" myself.

Re:What about the OS

[Anonymous Coward]

Linux will have a 64-bit x86-64 version available when Athlon 64 and Opteron are launched.. (see http://www.x86-64.org).

Microsoft and AMD have already demonstrated 64-bit Windows on Opteron. Supposedly Longhorn will be x86-64 enabled.. which is due out by end of 2003.

They still do

[ArchieBunker]

Talk to the right people and you'll find a beta of win2k server for the alpha cpu. At the time neither intel or AMD had a ready for prime time cpu and MS needed to keep a working 64 bit codebase.

Re:They still do

[Anonymous Coward]

Beta 2 for W2K/Alpha went out on MSDN, so it shouldn't be that hard to find.

Now, the legend is that the W2K build team still kicks out an Alpha version of the production package. That's what the 'right people' could get you.

AMD and RAID controllers

[Woodrose]

I hate to bring up the old question of instruction set compatibility (I'm wearing my nomex undies in any event ;-/ ) but a friend of mine owns/runs a small vertical market ISP and says he couldn't get his hw raid-5 controller drivers to load on servers using AMD chips. He tried several variations of controllers and a couple of different AMD servers. No problems doing it with P3's. I didn't think to ask him which versions. Hardware was all new (report was a good few months ago though) and the guy is not entirely clue-proof. Has anyone else seen this behaviour with AMD processors? Can someone give me a bit of confidence that these chips are really, truly compatible? I don't want to give my friends bad advice.

Re:AMD and RAID controllers

[cerberusti]

Actually current AMD processors have fewer errata than Intel ones, which is kind of ironic. The problems with peripherals are generally caused by the motherboard chipset, not the processor. Most compatibility problems are a result of VIA chipsets parking the PCI bus on the last device used, which is different than the way Intel chipsets do this (but is still perfectly valid.) Any device which does not operate correctly with this is actually in violation of the PCI spec, as it is perfectly legal to park it wherever they want. I would not trust a raid controller that does not even get this right... Some motherboards have an option to park the bus in an Intel compatible way, which would clear up that issue.

If not, it will be the first time.

[DaveAtFraud]

Every time the processor bus width has gone up (8 -> 16 -> 32), applications have expanded to "fill" the newly available bandwidth. A few data points:

1) I remember PC Magazine predicting that 80286 systems would only be needed as servers.
2) I get by with a 733 MHz system at work but the only system I can't saturate at home is my dual Athlon box by "just" playing a game. The more realistic and "imersive" the game, the more bandwidth it takes.
3) I correspond with several people who do digital video editing and they can swamp their dual CPU systems.

Nope, if that were true...

[raehl]

Alphas might have had a chance, being 2^32 times as good as their 32 bit competition.

DRM features?

[vandan]

It's nice that AMD are benefitting from open source developers such as those as http://www.x86-64.org
But keep in mind that AMD have stepped forward (with Intel) and said they will be planting DRM features in their products to satisfy M$'s trusted computing push. And while initially you will be able to turn them off, soon the US and their states such as England and Australia will pass laws to make such consumer disobedience illegal.

Re:DRM features?

[AbRASiON]

Damn I really, really REALLY wish you were wrong,... but that topic isn't reall the point of this discussion now is it?

Damn palladium,.... microsoft and the whole world domination thing might just come to fruition (sp?)

Re: Skip to the last page for the most interesting

[stratjakt]

How about.. Hey Ruiz, see YOU in court.

Unless he has proof, that would be libel, no? Of course not - in his world, when a competitor gets more business by offering a substantial discount that you cant match - thats bribery. When you do it, that's good business.

Maybe Gateway cant deal with the support cost of overheated (and dead) Athlons? Or maybe too many processor cores are getting crushed when putting on the heatsinks? Maybe they just DONT SELL to gateway's customers. I can think of a lot of reasons not to want to stock AMDs CPUs that don't involve some giant corporate conspiracy.

Now,

AMD has their thumbs just as deep in MSFTs Palladium pie, they just have smaller thumbs and leave smaller prints.

Why do you assume because a company is struggling in a market makes them somehow kinder and fuzzier?

What makes you think AMD gives a flying fuck about open source? Are they a non-profit orginization now?

If tomorrow, AMD magically had 80% (desktop) market share, and Intel 20%, would the world be a better place?

If you really wanted to make a difference, you'd boycott anything and everything digital, DVDs, CDs, video games, computers, and make a statement that you don't like the way the industry is treating us with DRM et al.

Anything short of that is not only pointless - it's hypocritical. Hear yourself saying "Intel is an evil company! Buy AMD!". It's moronic.

Moore's law...

[kilonad]

A hundred years? What, will Moore's law slow down by a factor of two? 32 orders of magnitude gives us 32 doublings of current technology, or 32 cycles of Moore's law... at the current rate of roughly 18 months, that gives us 32 cycles * 1.5 years/cycle = 48 years. So it should hold us over til around 2050, although I seriously doubt that anyone (personal users, the fed's databases will need all the space they can get by then) will be using up all that storage. Unless we start talking about gigapixel video or digital holographic video storage or something along those lines.

Re:Moore's law...

[Anonymous Coward]

Adding the third dimension to today's 2D image resolutions would eat 10 bits of address space alone.

Won't be fully optimized for a while

[Stonent1]

Just getting the latest gcc for 64bit amd and doing a make world isn't going to do that much for you. Real speedups come in non portable code, which around here...seems to be taboo.

the real question

[pangu]

I think the real question here is going to be what's the 64 bit line in the old saying that ends with:
"a 2 bit company that can't stand 1 bit of competition"...

Re:the real question

[Anonymous Coward]

Oh, you mean Windows will be a 64 bit rewrite of a 32 bit extensions to a 16 bit patch for an 8 bit operating system originally written for a 4 bit cpu by a 2 bit company that can't stand 1 bit of competition?

Great...

[xenocide2]

so now the registers will be named like "weax" and "xfbc"?

Re:Great...

[be-fan]

It's RAX, RBX, etc. Don't know what the 'r' stands for, but it makes for the funny RIP instruction pointer :)

Re:Great...

[Decimal]

It's RAX, RBX, etc. Don't know what the 'r' stands for, but it makes for the funny RIP instruction pointer :).

Considering that the 8 new 64-bit registers are labeled R7 - R15, I'm guessing that R stands for "Register". But what do I know?

Re: Skip to the last page for the most interesting

[Anonymous Coward]

Hey Intel, see you in court! Of course now that Intel is along with Microsoft backing a group to outlaw opensource in the government, I think its time for the opensource community to boycott Intel. Why should our money go to a company which is now attempting to hurt Linux and opensource? I know because these recent actions, I will NEVER buy Intel ever again!

Word up: intel doesn't care about you. The vast majority of it's profit comes from business sales, you are 10%. You do not matter.

But bribing? George W. would be proud: Intel is just succeeding the Republican way.

I know some engineers...

[inode_buddha]

who could really use this, I bet. Their typical job ATM has an array of at least 3000x2000 data points as input (starting the job) and the job requires differential calculus to the 12th decimal point for each possible combination... last I heard, their shiny! new! pentium 4s with WinXP were choking and gagging for days on end.... so like I said, bring it on, AMD.

-1: Funnyless

[Anonymous Coward]

Some people have the nerve to talk about the Alpha CPU... They just pull allegations out of the air and say the Alpha wasn't popular because it was too costly. Excuse me, friend? I own an Alpha and I traded them that useless golden calf for it! The Alpha is still ahead of it's time! When Intel produced the 233MHz Pentium, Alpha was ballet dancing to Alice In Chains at 600MHz. The Alpha was all about quality, the Pentium was all about affordability. Alpha was verry much backwards compatible with software just as well as X86's only in a sense that the Circuits simply got faster and only a few machine codes (PAL) were added and didn't overall destroy the backwards compatibility.

Then you people pull another nugget of Balmer's shit out of your Microsoft'en assholes saying, again, the Alpha is too expensive. Go visit eBay: you can still buy that 4 year old 164LX / 164UX / AlphaServer for... $500.00.

GADZOOKS! For being 4 years old, maybe you're right!

Maybe that explains why so many people are relunctant to upgrade...it's already faster than a Pentium 4 or Athlon rated at 2,000MHz. Alpha is such a well-designed architecture and Processor.

Someone's selling a CS20 oh wow!

Re:This is why I don't understand AMD's lament...

[symbolic]

Why does AMD have to compete with Intel on raw speed? Why does AMD need to have the absolute fastest chip available? Why not concentrate on providing customers with a low price and high performance instead- in other words, chips that may not be the absolute fastest, but are fast enough?

The reason my last two PCs have been AMD-based boxes is precisely because they didn't have Intel processors. I'd like to think I'll still have a choice a couple of years from now.

Here's where consumers need 64-bit CPU's

[MtViewGuy]

I think you'll be surprised where 64-bit CPU's may become useful consumers.

The first place where this will be useful is video editing. With the proliferation of MiniDV camcorders that have IEEE-1394 connections to desktop computers, many camcorder users are downloading video onto their computers for editing and creating home-made VideoCD or DVD-R discs. With 64-bit CPU processing we now can see the development much more sophisticated (yet easier to use) programs that make video editing and VideoCD or DVD-R disc creation almost a snap.

The second place this is useful is still image editing. With the proliferation of digital still cameras with USB ports people are doing more and more image processing of still images before printing out the pictures. With 64-bit CPU processing we can see image-editing tools that can do image processing that is far more sophisticated than what even Photoshop 7.0 can do today, yet would be easier to use than ever.

The final place is games. 64-bit processing makes it possible to do extremely sophisticated graphics effects in real time without over-reliance on an expensive high-end graphics card; a lot of games that need fast motion with complex backgrounds could benefit from going to 64-bit CPU processing.

Little tidbit for yall...

[Anonymous Coward]

This isn't meant as flamebait in that "anything more than 500 MHz is overkill"...

But I once read how the author of Counter-Strike (most played first person shooter, if not most played game PERIOD on the Internet), Gooseman, developed it on a celeron 400 MHz machine he self assembled...

Always thought that was pretty neat ;) 400 MHz celeron...it just goes to show you don't need the latest and greatest to make cool stuff (although at the time it was)

Its different for people in 3D Rendering stuff though... for my friend it takes his p4 2.0 GHz box HOURS to render stuff for him.

Re:Little tidbit for yall...

[AbRASiON]

You're obviously new to the gaming scene.

Counter strike came out not that much after half life (it is a MOD after all)

When half life came out a MONSTER pc was say a P3-500 or 600 (if I recall) and the majority of us "enthusiasts" were using Celeron 300a's @ 450mhz.

So of course he began developing it on a C400.
Also it's all about the video card in 3d games, not the cpu.

Famous last words

[LowellPorter]

640 TB should be enough for anybody

Didn't Bill Gates once said that 640K is all one would ever need? =)

Re:Famous last words

[Hobophile]

Link

No, he didn't.

Here's a suggestion. Whenever you come across a quote or fact or story so cute that you can't resist telling it to people, check with Google and see if it's been debunked.

The amount of myth and misinformation circulating through email and around the water cooler is staggering. Do your part to fight it.

The point

[Dave_bsr]

you, sir, just said it best of anyone so far, IMO.

"Once the processors are available, applications will be written to take advantage of the larger word sizes. There's no way to tell what will happen."

Exactly. What can't you do on your PC now, fast enough? full-motion, real-time, uber-quality video editing? Molecular/weather/beer modelling? Compression? Encryption? Random CPU usage for fun?

That's why we need better computers - because you can do nearly _anything_ on a PC, given enough processing power. That's why they are so beautiful: general purpose computing rocks. In ten years, I will look at my 32-bit Athlon, and sigh, and remember the days... :D

benchmarks?

[Dave_bsr]

ha. good point. benchmarks? we don't need no STINKIN benchmarks!

Re:Heat. tsarkon reports.

[Anonymous Coward]

Wrong. You are such a FUD/conjecture machine it's unreal.

The Intel shrinks of the P3 (.13) with the integrated heat spreader are quite cool to the touch under load. In fact, the FIRST CPU Intel shrunk was the Pentium III. You know, you can READ about CPUs. http://www.intel.com/intel/finance/pricelist/ , http://developer.intel.com/design/PentiumIII/specu pdt/ and for P4 lovers, the specification updates for the p4, http://developer.intel.com/design/Pentium4/specupd t/.

Now there is similar documentation available from AMD as well. Now instead of getting anecdotal evidence, (Which can show you clearly there are many holes in this guys Pentium 3 theory), you can see steppings, dissipations, packaging and thermal characteristics for CPUs, not some asshole's rendition of fact into half truths. You want to talk about heat, talk about dissipation first.

Intel could put out a 1.6GHz Pentium III tomorrow [The 1.4 is able to be clocked at that frequency]. They stopped the P3 at 1400 with 512cache for a reason - it started to compete with Pentium 4.

Do you expect anyone to believe you have ever read EETimes or have a sub to Microprocessor Report? I mean, you are poorly summarizing gamer sites here continually.

By the way, for anyone thinking of this guys trash, the Athlon shrinks and the hammer chips will run cooler. Maybe it wont be up to snuff with Intel.

Now the biggest problem with your filth is that Hammer specs aren't out. system integrators need thermal information to make designs. And when Dell and friends get the thermal data, along with everyone else, then you can talk about what kind of heat output AMD-64 will have.

Everything you say is conjecture, crap, fluff. You don't work for AMD, and judging from your intellect, if you did, they laid you off.

32 bits != 4 gig max

[cartman]

32 bit architectures are not limited to 4 gigabytes of memory. "32 bit processor" refers to the width of the DATA bus (and registers). It does not refer to the width of the address bus.

For example, the z80 and 6502 were 8-bit processors, but they supported more than 256 bytes of RAM (2^8 bytes). The 68000 and 80286 were 16-bit processors, but they supported more than 64k of RAM (2^16 bytes). That's because the 8-bit processors had 16-bit address busses, and the 16-bit processors often had 24-bit address busses.

The current pentium-4 Xeon chip supports 64 gig of RAM, despite being a 32-bit processor.

64-bit computing means that you can hold a 64-bit quantity (long int or double) in a register. Also, you can load, store, or perform arithmetic on such quantities using one instruction and often in one clock cycle.

This offers very few benefits for the end consumer. Mostly it's about perception: consumers will percieve that a 64-bit chip is twice as good as a 32-bit one.

Re:32 bits != 4 gig max

[mr_data_esq]

Mostly, I agree. In fact, I spend lots of time writing software for 8 and 16 bit machines, and I spend half that time turning single bits on and off.

One thing I'd like to point out, though: I've noticed that an awful lot of mathematics is being done using doubles (i.e., 64-bit floats) these days. It's partially laziness, but it's also really the case that 32-bit IEEE floats only give you 24 bits of accuracy. Doing math with doubles really cuts down on roundoff errors, so a lot of people switch to doubles and forget about it.

Re:32 bits != 4 gig max

[Bert64]

AFAIK the 68000 was a 32bit processor, with 24bit address bus and 16bit external bus. The later 68020 increased everything to 32bit.
However, the p4 actually has a 32bit address bus, with hacks to address 36bit space, but thats what it is.. a hack, the extra addressspace is not directly available to apps. There is also likely to be a performance hit when using these hacks..

Re:32 bits != 4 gig max

[rweir]

The current pentium-4 Xeon chip supports 64 gig of RAM, despite being a 32-bit processor.

Yes, that's true, but it's horribly hacky. Addressing your RAM in 4gb segments? It's enough to make any old-skool DOS coder cry.

Re:32 bits != 4 gig max

[Ninja Programmer]

• 32 bit architectures are not limited to 4 gigabytes of memory. "32 bit processor" refers to the width of the DATA bus (and registers). It does not refer to the width of the address bus.

This is a marketing term, not a technical one. And as far as the current 32 bit processors are concerned, you have it exactly backwards. x86's today have SSE registers which can hold 64 bit integers. I think the PPC's AltiVec registers can also hold 64bit integers. However, neither processor can access a 64 bit address space.

But even the marketeers knows that you don't call these processors "64bit". With the notable exception of Nintendo, the only time a processor is labelled "64 bits" is when its *address* space is 64 bits, not just its registers/ALUs.

Re:32 bits != 4 gig max

[Anonymous Coward]

By that definition, you have just upgraded my 8bit Commodore 64 to a 16bit CPU.

And the Pentium to some weird size like 36 or 48 bits.

It is not the size of the address bus that counts, it is the size of the *general purpose* registers. I.e. AX, BX,... on Intel, not the special-purpose registers in the FPU (80387), MMX, SSE, Altivec, GPU.

Re:32 bits != 4 gig max

[cartman]

What you said is incorrect. "32 bit" has ALWAYS referred to the width of the data bus/registers.

SSE2 and AltiVec have 128-bit registers. This is so they can hold 4 32-bit quantities (the point is to be able to do 32-bit operations on 4 values at a time). It is still a 32-bit processor.

If the number of bits described the address bus width, then x86 is 36-bit, DEC Alpha is 36-bit (NOT 64-bit), 6502 is 16-bit, 68000 is 24-bit, 80286 is 24-bit, etc. This is clearly not the case. When was the last time you heard anyone refer to a Xeon as a 36-bit processor, or a 68000 as a 24-bit processor?

Re:Atoms are Small

[Anonymous Coward]

64-bit really should be enough for a hundred years. If you could turn every atom of the earth into a bit of storage, 64 bit addressing would still be enough to give every byte a unique address. (or something like that.)

Not quite, 8*2^64 silicon atoms weigh about 7 milligrams. The earth has about 2^166.4 atoms.

P4 no longer cooler operating than Athlon

[Brian Stretch]

The new Thoroughbred Revision B Athlons (XP 2400+ and higher) made a significant drop in power consumption (1.65V core), while the 3GHz P4 guzzles more electrons than any Athlon (have you seen the heatsink Intel bundles with that thing?!). The Hammer series uses Silicon-On-Insulator technology to keep power consumption (heat) down, to the point that the larger Hammer core consumes about the same amount of power as the TBred RevB. AMD is gunning for the high-density rackmount market with the Opteron where efficient power use is critical. They'll get it too.

I have a dual CPU Athlon 2400+ box, 2GHz each, using Thermalright SLK800 heatsinks and 80mm adjustable fans set to 2500RPM. My temps are 41C/43C/42C (case/CPU1/CPU2) at the moment with about 25% CPU utilization. Power consumption (as measured by my UPS load monitor) is the same as the dual Athlon 1800+ chips (1.53GHz) the new CPUs replaced.

FUD disguised as a technical comment.

[Ninja Programmer]

• 1. All addresses being 64-bits.
  For #1, realize that this is going to greatly increase the data size of many applications. The larger the data size, the higher the chance of cache misses. In general, this is a loss, not a win.

This is incorrect. The Hammer "long mode" uses 32 bits as the default data size. 64 bits are only used for pointers and explicitely overridden 64 bit operands. I.e., you still have to declare "long long" or "int64" or whatever, in your languages to access those 64 bits. All your old 32-bit data still occupies the same space.

Furthermore, measurements by AMD indicate that op-code size did not increase with the expanded instructions, but actual *decreased* because the additional registers decreased the typical amount of spill/fill code emitted.

Therefore there is no additional cache pressure. The "code bloat" problem remains solely in the hands of the software developer, and is *NOT* worsened in any way by hammer.

• 2. All internal integer registers being 64-bits.
  For #2, realize that some integer operations are O(N) where N is the number of bits involved. 64-bit multiplication and division are slowerthan the same 32-bit operations. Period.

This is also incorrect. There are numerous well known techniques used in ALU design that makes precious few operations "O(bits)". Again, AMD specifically targetted this. For example: the 64-bit integer multiply in hammer is *FASTER* (per clock) than the 32-bit integer multiply in either the Athlon or Pentium 4.

The reason AMD is able to do this is because arithmetic and logic operations can largely be implemented in a "more gates for more speed" fashion. They are closer to O(ln(N)) than O(N). But at this level of circuit design, you don't necessarily think in those terms (since N is constant, everything just looks like O(1)) -- these high speed circuit designers worry about other technical things like "latch speed".

The 64 bit integer divide may be a little slower, however, again you need to explicitely use 64 bit ints in your software, and division is a comparatively uncommon operation.

• The gain with 64-bit processors is one of address space and nothing more.

This is the largest gain (big DB people will be very happy with it) but it certainly is not the only gain. Remember that there are now twice as many SSE registers. This opens up some performance possibilities for multimedia applications.

Although I don't know that its related to SSE, it should be pointed out that EPIC (as in the video game company) has ported the Unreal engine to x86-64! Like most people, I was quite surprised that they did this, however, they apparently found doing it to be worthwhile.

Do not underestimate the upside of going to 64 bits in the way that AMD has done it. They have literally made it a no-lose scenario -- that alone should spur (mostly new) application developer interest.

Re:FUD disguised as a technical comment.

[captaineo]

Thanks for the clarification. Just two minor comments. First, on "long" mode - it is my understanding that 64-bit Windows will use such a standard, since far too much Windows code depends on "long" being 32 bits. However, on UNIX style systems like Linux, "long" will be 64 bits (as it is on most other 64-bit systems).

I disagree that the move to 64 bits will not affect cache pressure. I can think of many applications where the data structures are mostly pointers (e.g. compilers or XML processors). In such cases, working sets could almost double as pointers become 64 bits.

Re:FUD disguised as a technical comment.

[Anonymous Coward]

There are numerous well known techniques used in ALU design that makes precious few operations

This is also why Hammer tips the scales at 100 million transistors...

Re:FUD disguised as a technical comment.

[Junks Jerzey]

This is incorrect. The Hammer "long mode" uses 32 bits as the default data size. 64 bits are only used for pointers and explicitely overridden 64 bit operands.

Now isn't that *exactly* what I said, that all addresses are now 64-bits? I didn't say anything about integers stored in memory.

This is also incorrect. There are numerous well known techniques used in ALU design that makes precious few operations "O(bits)".

Yes, and those precious few include multiplication and division. As another poster pointer out, you can get these down to O(log N) if you trade space for speed. But if you get a 64-bit divide executing faster than a 32-bit divide, then something is seriously wrong. That must be one really awful 32-bit divide.

In any case, there's no doubling in speed to be had by going from 32 to 64 bits. That some people are quibbling over a few cycles in a divide unit is pretty sad overall.

Oh yeah!!

[Penguin Follower]

4 GB ram drives!! :) (just kidding)

Minor nitpick

[Inthewire]

I'm basing this on my ignorance and faulty memory, but CISC wasn't developed to mimic HLLs. C was designed by Dennis Ritchie to mimic available (PDP-11) systems.

Re:Minor nitpick

[ameoba]

Perhaps I worded that poorly. Rather than the the ISAs growing to resemble existing HLLs, many CISC architectures have complex instructions giving a higher-level idea of what the CPU is doing. The x86 even has text-manipulation opcodes, and the VAX (commonly refered to as the epitome of CISC computing) has 8 different ways to interpret the value of a register.

RISC machines are intended to be programmed by compilers; CISC are designed to make life easier for programmers.

Re:This is why I don't understand AMD's lament...

[Anonymous Coward]

If it ever gets to the point where there is no choice, the Government will step in with a nice anti-trust suit.

Do it yourself Re:Heat.

[draziw]

If you don't know how to solder, learn to. Solder two 1 Watt 50 ohm resisters in parallel on the + lead of your fans.

You will get a very nice drop in noise. (And CFM) Add at least one fan to counteract the drop in CFM. You can also replace your louder fans with ones that start out quieter. Take a look at Sharka Computers. You can also order fan-mates if you don't want to solder, and each one can control a few fans.

Good luck!

Ryan

We might need more than 64 bits

[Luminous Coward]

Some people are claiming that 64 bits is all we will ever need. 64 bits will "only" encode approximately 2*10^19 values.

Imagine I want to give each byte of storage available in a cluster his own unique address. I could then simply write MOV RAX, [RBX] and have the OS figure out I want 8 bytes located on a completely different computer's disk or memory (Single Address Space Operating Systems explore this possibility and they actually bank on a sparse address space).

Consider a large corporation with 10^6 devices and each device with 100 GB or 10^11 bytes available. 10^17 is close to the 64-bit limit, don't you think?

Hard Numbers: Build Times!

[Anonymous Coward]

Ok all I have to say is that in some environments faster means money. Take a code shop with 50 developers at a modest $60000/year each. A full build of our system takes a P-3 600 Dell Optiplex about 1 hour 20 minutes. The new workhorse 2.8 gig machines we just bought: 14 minuts flat. total cost of each machine $800. Estimate that each developer does 1.5 builds a day (easily). lets say 48 work weeks a year, 5 days a week... thats 450 build hours a year per developer... with the new machines... about 72 build hours per year... estimated raw (not real) savings: almost $600,000 per year.... ok lets say that in real savings you only get a 20% return on those extra work hours: you've saved almost $120,000 on an investment of $40,000 for a net of $80,000.

Not to mention your developers are a heck of a lot happier! (you try waiting an hour and 20 minutes for a build!)

The benefit is much higher near release time when there is a time crunch, especially with varients of a project, and official builds. At the end of a project each varient must be built, some of our products have 20! or more varients, at about an hour 20 build time each. Trust me, someone will find plents of "desktop level" use for this horsepower! (I sure wish I could ditch that 600mhz P-III)

There is no "desktop" market for 64 bit CPUs

[ppetrakis]

I don't have to read the article. I've been working with Alphas all my life. There is nothing for 99.9% of the applications you use everyday that could benefit from running itself in a 64 bit address space. Unless you get a signifigant performace boost from the move (like Alpha in it's heyday) it isn't worth the effort.

If you find you need that sort of mega addressing the chances are the app you need already runs on 64 bit Solaris. After that point it's up to the vendor (Think Avanti Corp /Apollo) Wheither it's worth their while. Remember, You need their application. Unless your app is home grown or you have some signifigant pull with a vendor the port isn't going to happen.

The desktop is an afterthought. This chip was designed to be sold in quanties of 8 and higher in single large servers. Once they cut into that market the economies of scale just happen to make it cheap enough for the desktop market to pick it up. They have a much better chance at getting it down with their builtin backwards compatibility and keeping costs down. Alpha never hit that "sweet spot" for the volume to really bring down the price..

Now, Don't think Intel is going to sit on its hands while AMD eats their lunch. They're more likely to drop an Itanium instruction decoder into an Alpha EV7 core and push that than follow with an x86-64 processor line. Itanium is just to big and costs too much to at this stage of development to make inroads fast enough stop AMD in gaining marketshare but more importantly, mindshare. Intel would never take up x86-64, Doing so admits defeat to the industry i.e. You're not the leader anymore.

So to sum it up, Intel will either:

1. release Itanium and we all find out it isn't as slow as everyone claims it is or as expensive
2. See Alpha/Itanium hybrid core above
3. They bring back Alpha (maybe not by name) and put it under a modern process. Expect atleast x1.5 current clock speeds and Alpha's can milk rambus for all it's worth

2 and 3 are much more likely than one, You know which one I'd rather see happen :).

Either way it'll be a boon for the OS community and certainly make our (The Alpha community) lives easier. The way I see it, even if hammer is moderatly successfull. You guys will 'clean' most of the popular soucecode out there to be 64 bit clean, reducing our matainence work by like 80%. The only thing we'll have to worry about is firmware, toolchain, libc, Xwindows, and kernel. So please buy a *hammer and learn the joys of porting to 64 bits. If it proves too painfull, please see the ld manpage for the "-taso" flag :).

Peter

Re:There is no "desktop" market for 64 bit CPUs

[bstadil]

Intel would never take up x86-64, Doing so admits defeat to the industry i.e.

Intel is a very pragmatic company. If they felt at some point that they needed to introduce x86-64 they will do so. They have enough marketing clout to pull it off.

Look at the way they are handling the Itanic. It is being positioned narrower and narrower, it will never die just "fade away".

Re:There is no "desktop" market for 64 bit CPUs

[Anonymous Coward]

There is nothing for 99.9% of the applications you use everyday

You fail to realize that once the 64 bit CPU becomes the new standard, applications will be developed that DO need and take advantage of it. That is the point of progress. Your statement is similar to saying there is nothing for the next generation video card, since none of the games today can use the new features on the chip. The point being, once the card is out, games then come out that do take advantage of the new features. The same principle applies here.

Re:There is no "desktop" market for 64 bit CPUs

[Anonymous Coward]

There is nothing for 99.9% of the applications you use everyday that could benefit from running itself in a 64 bit address space. Unless you get a signifigant performace boost from the move (like Alpha in it's heyday) it isn't worth the effort.

Technically, I would probably agree with you. But when you look at the whole chip, the 64-bit AMD CPUs due bring something to the market. 1) Additional 8 GP registers 2) Additional 8 SIMD registers + SSE2 support 3) Removal of (basically) the entire northbridge = less reliance of chipset makers = more consistent performance The average office/web user won't gain much, but FPU intensive apps like media and games will (continue) to drive the market.

So to sum it up, Intel will either: 1. release Itanium and we all find out it isn't as slow as everyone claims it is or as expensive 2. See Alpha/Itanium hybrid core above 3. They bring back Alpha (maybe not by name) and put it under a modern process. Expect atleast x1.5 current clock speeds and Alpha's can milk rambus for all it's worth 2 and 3 are much more likely than one, You know which one I'd rather see happen :).

Option 4) The upcoming Intel Prescott core will have 64-bit registers on silicon, just disabled, so it wouldn't be too much of a stretch for Intel to move to x86-64.

Re:There is no "desktop" market for 64 bit CPUs

[Decimal]

I don't have to read the article. I've been working with Alphas all my life. There is nothing for 99.9% of the applications you use everyday that could benefit from running itself in a 64 bit address space.

After looking at your title and seeing no relation to your first paragraph, I know I wouldn't have to read the rest of your post and still know exactly where you went wrong. The "market" for a computer is not necessarily defined by what new applications it can offer. It can be defined by Joe Average coming home to his house carrying a huge box, telling his wife "You've gotta check this out. Its got sixty-four bits!"

(Chances are he's never worked with a "computer" in his life, and thinks he'll have to assemble all 64 pieces manually.)

Re:There is no "desktop" market for 64 bit CPUs

[ppetrakis]

I understand and agree with that. Remember back when AMD first release their own competing instruction set to MVI? You could have 3dnow etc etc but if the application didn't use it, it didnt matter. In the meanwhile, that chip runs windows just fine. It's the same idea but this time it's 64 bits. Though this time it's not as simple as optimizing program X to use 3dnow, You're porting the whole code base to 64 bits.

So yeah you have some dudes going out and buying this cuz "wow it's 64 bits". Then they'll find that nothing they use takes advantage of it. You're not gonna toss your chip are the window? Hell no. The chip is still fast as hell standing on it's own and it's not like you have an option to buy a 32 bit AMD cpu anymore so you're basicly getting it for 'free'. It'll be up to OS and application vendors to decide when the market has enough saturation and they can make some more money, will they then port their application to 64 bits.

The point of my post was to make the readers here aware that moving to 64 bits is not a trivial matter. You guys won't see major application ports for years though in the meantime you have a blazingly fast CPU that does what you need it to do -now-. That offers those who need that capibility to use it for their own applications.

Never did I say AMD will lose, I was looking at the big picture. AMD will probably snatch up the low and midrange markets immediatly. When OEMs/VARs start the to see the $$$ of moving to AMD completly. The only thing that will bind them is the cost of supporting existing Intel PCs. Think spare parts & warrenty replacement for somebody like HP or IBM.

Like I said, The desktop is an afterthought. It just so happens that the price falls into that range.

Peter

Re:There is no "desktop" market for 64 bit CPUs

[ppetrakis]

Sure they have the clout but do they have the money anymore to make that kind of move? Also they have some big companies heavily invested in Itanic -now-. If Intel nix'd Itanic and went with x86-64 or Alpha would you still buy from Intel being after being their test mule?

In the eyes of the desktop and midrange market it would just fade away but for the big boys who spend countless millions on these installation would throw a fit. They could lose those big customers forever, Why stick with a company that can't deliver? Assuming hammer is delivered and is a success, AMD's track record would be much better off than Intels. If you where to pick a company on "can you get the job done", AMD would win.

If Intel goes with x86-64, Mark my words; It's the beginning of the end for Intel. We're going to trade one chipzilla for another.

Peter

Re:There is no "desktop" market for 64 bit CPUs

[ppetrakis]

Oh, I realized it. I just failed to mention it :). I've been doing this for along time so please forgive me if I forget to mention a nuance or two :).

http://slashdot.org/comments.pl?sid=46635&cid=4802 010

This time, It's not as straight forward as AGP 4x or something. There are huge costs in porting existing code bases to take advantage of 64 bit. Newly designed applications will be fine (see previous post).

I'll reiterate once again. The desktop is an afterthough, You get the extenstion for free. It's in the servers where the money is to be made.

Peter

Re:There is no "desktop" market for 64 bit CPUs

[ppetrakis]

With regards to your first point.

http://slashdot.org/comments.pl?sid=46635&cid=4802 010

With regards to option 4:

http://slashdot.org/comments.pl?sid=46635&cid=48 02 128

With an emphasis on the last line.
"If Intel goes with x86-64, Mark my words; It's the beginning of the end for Intel. We're going to trade one chipzilla for another"

Peter

Windows costs $2000 per 8-cpu box

[yerricde]

I.e. 8-way becomes a PCB-design issue, not a chipset issue.

And an OS support issue. Of all Microsoft operating systems, only Microsoft Windows 2000 Advanced Server and Microsoft Windows 2000 Datacenter Server support 8 CPUs, Don't expect 8-way Windows to cost less than $2000 per copy, assuming that the OEM price is half the retail price of Windows 2000 Advanced Server with 25 CALs.

Re:Windows costs $2000 per 8-cpu box

[Anonymous Coward]

When an 8-CPU box costs at least $50K, does one really care?

Besides, your RDBMS licences will be much greater than your OS licence.

I like the word plot...

[Kjella]

...because I'm fairly sure that it's not that big a deal as they make it out to. Hopefully it won't be as bad as the Pentium Pro, but I'm guessing it won't run faster on 32bit code, which is what 99% of the programs will be at launch.

Look at it more as their version of the GHz game. "Our processor runs at 3GHz, yours at only 2GHz" "Yes, but we have 64bit registers, yours only has 32bit" Neither is a good measure of actual capabilities. But I'm sure it'll help AMDs marketing department.

Kjella

Thank god for 'em. I landed a job because of it!

[iamwoodyjones]

I'm working on several projects right now dealing with these problems. The thing is that you're right. These misnomers are hacky things done by not-so-good programmers.

Unfortuanetly explain that to a FORTRAN program that was written in the early 70s and didn't have a pointer. It had to use an integer.

Yes, a lot of old stuff is shit and now programmers who thought, "If it's not broke don't fix it" are getting burned.

Compiling 32-bit != 64-bit app

[justanumber]

Also, "porting" and "refactoring" aren't necessary. Just recompile. CFLAGS="-m64". What a rare insight. Ok breathe deeply and concentrate. Just because you compile this with a 64-bit compiler does not make it a 64-bit OPTIMIZED, REFACTORED application. Get this straight. It makes a 64-bit CROSS COMPILE of the 32-bit applciation you had before. This is why AMD is doing this: http://www.amdzone.com/releaseview.cfm?ReleaseID=1 035 And linux is continuing to do this: http://216.239.51.100/search?q=cache:2Sh4LRuHKFEC: www.gnumonks.org/ftp/pub/congress-talks/ols1999/ia 64-slides.pdf+porting+linux+64&hl=en&ie=UT F-8 Even though I can already do this: CFLAGS="-m64" to this: http://www.apache.org Clear? Clear.

Re:Compiling 32-bit != 64-bit app

[platypus]

Well, that linux can't only be recompiled for a new platform in nearly every case is clear, since parts of it are coded in assembly, it needs drivers for the mb-chipsets and whatnot. This is where the arch/ subdir from the kernel tree comes into play.

For userland apps the story is completely different, the needed changes nearly all are cleanups, where the code wasn't 64bit clean.
IOW, the statement that you can do CFLAGS="-m64" on any app isn't true.

Re:Compiling 32-bit != 64-bit app

[Eric Smith]

You still haven't explained how refactoring enters into it at all. There's nothing wrong with refactoring, and it can improve 32-bit applications without needing a 64-bit CPU. But it is neither necessary nor sufficient to take advantage of a 64-bit CPU.

It makes a 64-bit CROSS COMPILE of the 32-bit applciation you had before.

Or you can compile natively on the 64-bit machine. It shouldn't make any difference.

It is true that some applications will break if you simply try to compile them on a 64-bit platform. This is because those applications violate the C standard, typically assuming that sizeof(int)==sizeof(void *). Applications that make such assumptions aren't necessarily portable to other 32-bit platforms, let alone 64-bit. If you write your application badly to start with, it's not a surprise if it's harder to port. But if you write it well, porting to Hammer (or IA64, or Alpha) should be a piece of cake.

You can refactor til the cows come home, but that won't by itself isn't going to help with porting a poorly written application.

Re: Skip to the last page for the most interesting

[NSParadox]

Got a link to support your claims about Intel trying to outlaw open source? Or heck, even Microsoft? Microsoft is anti-GPL, but they are not trying to outlaw it, and they fully support BSD and other licenses where they can use source.

A babel fish that works?

[kaxman]

There's no way to tell what will happen.

You, sir dentarthurdent, are correct.

Re: Skip to the last page for the most interesting

[mchappee]

> I know because these recent actions,
> I will NEVER buy Intel ever again!

AMD's Sanders testifies on Microsoft's behalf as 'favor' to Gates

I will NEVER buy AMD again. Who's the bigger fool, the fool, or the fool who follows him?

Matthew

Re: Skip to the last page for the most interesting

[Dun Malg]

Unless he has proof, that would be libel, no? Of course not - in his world, when a competitor gets more business by offering a substantial discount that you cant match - thats bribery. When you do it, that's good business.

Maybe, maybe not. When Standard Oil undercut all its competitors by pricing its products BELOW production costs in order to drive them to bankruptcy and buy them out, that was ruled A Bad Thing and led to SO being broken up. There is a point where offering "special deals" is considered anti-competitive. If Standard Oil got nailed simply for offering product for too low a price, it's not unreasonable that Intel should likewise be nailed for offering product for a super low price, but only to companies that don't buy from Intel competitors. That's kind of shady territory there.

For example: BobComp buys Intel and AMD CPUs, so they get P4s for $35 each. JoeComp buys only Intel so they get the "deal" of $30 each. If BobComp buys 1000 CPUs a year from Intel and JoeComp buys only 500, then it's clearly not a "bulk discount".; it's a "helping us ace out the competition" discount. Now if $30 represents a significant loss of profit margin over $35 for Intel, then I'd say Intel is edging into some pretty anti-competitive territory.

is M$ quiet about anything?

[twitter]

MS have been quietly getting ready for 64 bit for at least 2 years; they've been shipping a 64 bit SDK on my MSDN disks for over a year. There are 64 bit NVidia drivers for WinXP-64. What makes you think MS isn't already there?

Spare me the smoke and vapor. Don't you remember the sad story of Mica, errr, NT on Alpha? Loudly proclaimed, quietly killed, that's why I think they are not there. If you consider the number of bugs and holes in 32bit M$ work, you might conclude they never arived anywhere.

In the mean time, you can get Linux and BSD on Alpha and other 64 bit platoforms:

• Debian
  
• Red Hat
  
• BSDs, free, net, open
  
• and plenty of others less well known.
  

Oh, it hurts so much to remember and think!

Re:is M$ quiet about anything?

[JKR]

Oh, it hurts so much to remember and think!

No, it hurts so much to pay a development team to support a platform returning a tiny fraction of your overall OS revenue. MS is a business, not a charity; if your platform can't compete on price with lowest-common-denominator hardware, they're not going to spend vast amounts of time and money supporting it. If they don't get a return on XP64 it'll get dropped, plain and simple. They've done exactly the same thing with CE - SH3 and MIPS support has been dropped, because ARM is the perceived best bet.

Remember the 80/20 rule - chasing the last 20% of the market costs you 80% of the effort.

Jon

Impatiently Waiting

[eigerface]

I don't have time to wade through another pre-release article on Hammer.
Have they mentioned anything even remotely close to a release date.

It's hard to rate a phantom.

Well

[autopr0n]

However, The GPU and processor are totally different.

Yes, thats correct. But the Nintendo 64 had a 64 bit processor (and a 128 bit graphics bus)

The "bitness" of a micro-architecture

[Luminous Coward]

I have to share this insightful comment I read on Usenet 3 years ago:

The "bit width" of a CPU is not strictly defined by a single architectural
attribute. Several candidates for a "normative" bit width exist:

- word width of the general purpose registers

- width of internal data paths

- width of external data paths

- width of the ALU

- width of the architected address range

There are probably more...

Back in the days of the 8 bit processors, ALUs were 8 bit wide, but address
range was already 16 bit.

In the age of 16 bit processors, registers and ALUs were 16 bits wide, but
often there were more than 16 address bits. Segmented addressing was needed
to make use of more than 64 KB for a single process.

When the first 32 bit CPUs appeared, they had 32 bit wide general purpose
registers and 32 bits of architected address space. But for example the 68000
had only a 16 bit ALU and its data bus was only 16 bits wide. Of the address
bits, only 24 were externally visible on pins.

Nowadays, with "64 bit CPUs" a reality for high-end computers, the address
width is the important criterion. Only a true 64 bit machine can linearly
address more than 4 GB for each running process. And when you handle pointer
variables that are 64 bits wide, it makes a lot of sense to have 64 bit wide
registers, a 64 bit ALU and 64 bit wide internal data paths. All current
64 bit CPUs that I know of meet this definition of "64 bit".

Internal bus widths tend to be wider (think of the 256 bit wide backside L2
bus of Coppermine or the G5), and registers have been wider than the "bitness"
ever since FPUs have moved on-chip (you don't even need to consider AltiVec or
SSE). External buses are sometimes narrower (to save some pins and a lot of
bucks on packaging) and sometimes wider (to better feed the new and fast CPU
cores from the same old memory chips).

So, by all intents and purposes, the x86 architecture was 16 bit until and
including the 286, and is 32 bit from the 386 onwards. AMD's K8 will probably
extend it to 64 bits. The P6 core is 32 bits, but it has some extensions to
enable it to address 64 GB of physical RAM. But every single process can only
address 4 GB directly, since pointers are still 32 bits wide. AFAIK K7 and P7
also have these extensions, but are still 32 bit cores.

BTW, the G5 is also rumored to be able to address 64 GB of physical RAM.
There are four unused bits in each of the "segment registers" which could be
used by the OS to select one of sixteen banks of 4 GB each. But processes
would still be limited to 4 GB of directly addressable memory.

Holger Bettag

because!

[twitter]

At first they will be expensive, then they will be in the $599 desktops. Why wouldn't you use them?

Could it be for the same reason people who have "high speed internet connections" still place charge by the minute long distance calls instead of using voice over IP, or crawl to a company like Yahoo to be a middleman? Look to Redmond, there you will see the answer. Anti-competitive, greedy stupid shit. Remember the 64 bit Alpha? You would think it would be cheap to make by now, no? Sorry, that might hurt someone's bottom line.

In short, I'd love to use one but I'm not going to get my hopes up.

A really fast processor means ...

[wytcld]

that it will become trivial in terms of speed to compile your distribution (a la Gentoo) and so rpm- and deb-based distributions of pre-compiled (and thus not optimized to your specific CPU) programs will lose their current speed-of-installation advantage. Having really fast CPUs means that you can have an OS with major optimization to the specifics of those CPUs and associated hardware. This ease of customization might produce a situation parallel to the one in which the specialty steel business remains profitable while the old, monolithic producers require massive subsidies to survive. That is, it could hurt both Microsoft and Red Hat - while being very good for open source.

Audio applications dummy.

[Anonymous Coward]

I am anxiously waiting for 64bit processors to run my virtual studio. You need lotsa horses for Cubase SX + a bunch of virtual instruments.

Um...

[autopr0n]

They have this thing called "linux" now. You should look into it.

AMD isn't...

[jpmorgan]

AMD isn't up to 3ghz yet. Perhaps that's part of the reason why they're interested in becoming more independent of the desktop market?

Yamhill Performance

[Best_Username_Ever]

Intel will scramble to introduce a new chip Yamhill

Rumour has it that Intel's development of the rumoured Yamhill chip was abandoned because it couldn't match the performance of Opteron. I would take this with a grain of salt though, since neither chip exists in the wild, and only one officially exists even as vaporware. But hey it might be true..

Dangit.

[Dave_bsr]

Knew i forgot something. NO VIDEO CARD! ARGH! Still, you can get a decent one for under $85, keeping the price at $600. *shrug*. All Hail Newegg!

Re: Skip to the last page for the most interesting

[jpmorgan]

The flaw in your logic is that Intel's actually making a profit, while AMD is still, I believe, in the red. Seeing as how it tends to be difficult to turn a profit when your primary product is sold at a loss, I'll take a stab in the dark and guess that they're not actually selling any chips for below the production cost.

Also, don't forget that Intel's manafacturing technology is about three years ahead of AMD. Their production costs are half of AMD's per unit.

A big problem with AMD chips, and something that I suspect is a not insignificant factor when it comes to the big OEMs, is that AMD builds fragile chips! If I need to build and ship x thousands of computers per day, if half the chips I buy get cracked during installation, I'm effectively paying double the unit cost.

Re: bashing me

[Dave_bsr]

Well, while you're bashing me,

PC-Chips, afaik, is not ECS,and their Sis board is very, very good. Somehow I doubt it is illegal...being sold online and all. I've had no problems and it's been more stable than other, more expensive boards I've used. Yeah...I build systems, I'm not just making this up.

Liteon was reviewed by Linuxhardware.org, and it beat out several Plextor drives.Their products are solid and fast.

Chieftec, as far as I can tell, makes the cases that Alienware uses. Yeah. And the PS was by Foxconn, if you bothered to look.AMD approved.Where did you get 430 watts from?

So I don't know quite what you are talking about. I've built 2 systems essentially like the one above. I've checked up on the users, and both are completely satisfied, say the systems are stable and amazingly fast. You say 5400 RPMs is slow...ok, but where did you get any idea that I was talking about 5400 RPM? the one i listed was 7200 RPM.

One thing you missed, while you were bashing me - I forgot a video card. Buy a nice GeForce 2 or 3, they are nice...don't get an MX.

So, you can do it, you can build a nice, sweet system for about $600, if you do it yourself. It's much easier (and educational!) to do it that way. AMD's price/performance ration is quite nice, and I don't mind taking advantage of it.

And whatever burr got up your butt, I'm sorry. A little more background info, maybe?

Uh....

[jpmorgan]

Who moderated this up?

64-bit CPUs have been around for years. Linux and Windows both run fine on 64-bit platforms. There is more to computing than PCs and Macs.

Won't make any difference

[jpmorgan]

Uh, 64-bitness is due to the size of integers and the memory addressing that the chip supports.

Numerical approximation of differential equations is a floating point problem, and you wouldn't use fixed-point math (integers) to do it (unless you were nuts). The P4 (and the Athlons) support 80-bit extended double precision floating point math, and with SSE2, it's damn fast. If they want really insane performance, they should consider grabbing one of the Itanium2 systems from HP. Intel are floating point gods; wwhy do you think SGI is planning to migrate their supercomputer lines to IA64?

Re:Won't make any difference

[inode_buddha]

Ok, this is news to me, but then I never was a math major. Nor have I really followed chip devel since the P3, but the AMD stuff looks interesting. Unfortunately, hardware is not in their budget for the forseeable future; that was finalized at the beginning of the contract 10 years ago, as part of figuring the profit margin. Likewise, the software install (Win2k/XP Pro)is standardized by the corporate, and can't be touched.

I thought the 64-bit thing referred to the width of the data bus, and how fast you could keep the math co-processors fed. In this case, the floating-point is especially critical, right? So, I stand corrected it seems. If you know any ways around all that, I'll be glad to listen.

GAH!

[autopr0n]

Stop using these base ten exponents while talking about computers, you're making my brain hurt!

Re:GAH!

[Luminous Coward]

Stop using these base ten exponents while talking about computers, you're making my brain hurt!

I am always happy to oblige.

Consider a large corporation with 2^20 devices and each device with 128 GB or 2^37 bytes available. 2^57 is close to the 64-bit limit, don't you think? :)

Re: bashing me

[shepd]

>Well, while you're bashing me

Sorry man, didn't mean to get personal. Just stating my opinion on ECS trash, not you (I'm sure you've got a clue... not a lot of slashdotters don't).

>PC-Chips, afaik, is not ECS

Here's the truth table of who PC-Chips is. They like to use new names every few months because they get tarnished so fast. FYI: Amptron (a known PC-Chips company) makes the identical board, in brown rather than purple (or is blue ECS's colour now? I don't know -- PC-Chips is pathetic at naming their products). I know. I inspected them side-by-side very closely. Trace for trace, component for component, identical in every way but where the FCC logo was/wasn't silkscreened (probably fake, too).

You can find out more about the quality of PC-Chips products here.

BTW: I know the K7S5A is actually labelled K7S5A, but to help other slashdotters, if you come across the 95% of other PC-Chips products that don't have a brand name listed, or simply have a removeable sticker indicating whatever was the popular chipset at the moment, you can find your board here (maybe). PC-Chips is the only company I know of too embarassed to label their products.

This is the list of PC-Chips names:

PCChips = Ability = Alton = Amptron = Aristo = Asia Gate = Asiatech = Assa = Atrend = ECS = Elpina = Eurone = Fugu = Fugutech = Hi Sing = Houston = Hsing Tech = H Tech = Matsonic = Minstaple = PCWare = Pine = Protac = QDI = Warpspeed.

I'll find the Amptron part number for your board if you'd like (they have big ugly pictures on their site you can use for comparison)...

>Liteon was reviewed by Linuxhardware.org, and it beat out several Plextor drives.Their products are solid and fast.

Let's ask them again in 6 months... Rest assured, you can ask any retailer, there's no way a Liteon will beat a Plextor. That's like a Kia beating a Lexus. Maybe in MPG...

>And the PS was by Foxconn, if you bothered to look.AMD approved.Where did you get 430 watts from?

Sorry, since you didn't provide a link, I clicked on the first case that came up. This one. Which meets your stated requirements. You get the 430 watts by adding up the stated maximum supply power for each important rail (ignore the -ve voltages and the VSB voltages -- these really won't contriubute to the usable power amount supplied by the power supply). This is what came up when I searched for your terms. The picture clearly shows it includes an "Austin" power supply with the usual labelling issues that show up on cheap Asian imports. At least it isn't likely to explode like an L&C/Deer power supply (I made a good few bucks during the last big storm!)... :-)

>So I don't know quite what you are talking about.

Sorry, if you'd like to be more clear as to what I should search for, perhaps I'll take a look for you. Newegg's site is a PITA to link to.

>I've built 2 systems essentially like the one above.

I've built 20 or 30 PC-Chips based systems. Only about 10 got returned, for various issues (about 5 related to me, although I don't consider them my fault, since the boss asked for it... the store was a sweatshop and we went to extremes like bashing in the power supplies to fit parts in because we catered to cheapasses).

>You say 5400 RPMs is slow...ok, but where did you get any idea that I was talking about 5400 RPM? the one i listed was 7200 RPM.

You didn't specify the speed so I had to guess.

>One thing you missed, while you were bashing me - I forgot a video card. Buy a nice GeForce 2 or 3, they are nice...don't get an MX.

I forgot the K7S5A is one of the _extremely_ few PC-Chips boards that doesn't have some trash Trident or SiS video built in. Whoops on my part!

>So, you can do it, you can build a nice, sweet system for about $600, if you do it yourself. It's much easier (and educational!) to do it that way. AMD's price/performance ration is quite nice, and I don't mind taking advantage of it.

Except I'd never buy it... and no one else would after their first experience with PC Chips.

I can honestly back up the fact that PC Chips hacks their Pirated BIOSes to show fake CPU speeds and fake amounts of cache. And that's just the start...

The question is: What are the requirements of the user?

My last PC-Chips board corrupted my HDD. If you don't care to much about your data, and don't need long uptimes, PC-Chips is a fine product, and the price can't be beat! Another one only boots up with the video disconnected. At least it boots up (I must stop being so masochistic, but I just seem to end up with so much PC-Chips stuff for next to free).

The best are the older PC-Chips VIA C3 boards that refuse to work with WD 2.0 Gig HDDs (when you work with cheap stuff, you get old parts. That's the law of the land). BIOS locks up at some point. Didn't bother looking for an update, I'm used to most PC-Chips boards not having any.

(And yeah, Tom's Hardware gave the thumbs up to the K7S5A. Which is why I don't trust anything he says anymore. He's clearly not been "in the business" for very long at all).

Ok, that's my last PC-Chips rant for the year. I'll just write a journal entry and let all the system builders weep in their beers over the woe that is PC-Chips.

Re:This is why I don't understand AMD's lament...

[grape_soda]

the reason why they HAVE to have the speed war is the same reason someone else posted. no one looks at benchmarks while they are out shopping at best buy or wherever they go... mom and pop and the college freshman just look at the numbers.. a 2.7GHz P4 LOOKs better then 2.1GHz AMD and thats all that matters.. people dont know that a 2.1 AMD can spank an Intel in most "buisness application" benchmarks. no matter what type of advertising you do, no one will believe that bigger is not necessarily better.

Comparatively

[be-fan]

You mean comparably? It's note quite than/then, but it's pretty unidiomatic.

100 Years?

[Anonymous Coward]

2^64 is certainly enough for at least a hundred years

Well, let's see. Moore's law (which is actually an observation, not a law) states that IC density doubles every 12 to 18 months. It turns out that address space usage seems to follow this trend as well. Put another way, we need an additional bit in address space every 12 to 18 months.

Following this backward, we should find the dawn of computing (0 bit address space) between 32 and 48 years ago (32 times 12 to 18 months). Intel produced the 4004, the first microprocessor, in the early 1970's (about 32 years ago). IBM was just starting to introduce commercial computers in the mid 1950's (about 48 years ago).

Since most currently consider a 32 bit address space near "end of life", following this logic forward tells us that 64 bit address space (an additional 32 bits) would be near "end of life" in 32 to 48 years.

Will this happen? Will Moore's law continue? Hell, I don't know, but it gives us a number to work with, and the number is substantially less than 100 years.

I'm an unregistered coward; mod me down.

Yes....

[Inoshiro]

FPU units which can communicate faster over the bus, since they send 1 data word per clock tick, instead of requiring 2 ticks to send 1 dword.

Re:Yes....

[psamuels]

FPU units which can communicate faster over the bus, since they send 1 data word per clock tick, instead of requiring 2 ticks to send 1 word.

Where are you, in 1993? Even the crusty old Pentium has a 64-bit memory bus. Bits of address space / register size are only loosely correlated with memory bandwidth. Granted, the Hammer will have a buttload of both, but keep the issues straight, will ya?

Note that in most circumstances it is just stupid to use 64-bit registers. If your application runs fine with 32-bit quantities, you can save a lot of memory bandwidth - not to mention a lot of memory - by using 32-bit registers.

What's cool about the Hammer is

• 64-bit integer registers for those (few, in the home market at least) applications that really need them
• massive memory and SMP bandwidth for everyone else

not necessarily in that order. (Indeed, "everyone else" will probably only have 1 cpu per box, for that matter.)

Hey, great.

[Inoshiro]

The 286 could address 16 megs of RAM, too. In real mode, you still had to deal with a 64k segment view on the world.

The PPro has, at best, a 4gb segment view of the world. This is because it's still very 32-bit. Unless it's all 64-bit, it's of limited use.

Re:Hey, great.

[G-funk]

The PPro has, at best, a 4gb segment view of the world - Not if you have 2mb or 4mb paging turned on (as opposed to the default 4kb paging)

Re:Hard Numbers: Build Times!

[Jeff DeMaagd]

I do believe that there are people that need the horsepower. But there are probably 10 people that don't for the one that does. Animators, video people, developers, engineers and scientists probably do. People running Office anything or some web client probably do not.

Oh my!

[Inoshiro]

That is the funniest joke I've heard in a while :) I hated those gray boxes. ;)

It's still a hack.

[Inoshiro]

Until I can have an entire virtual space that allows a complete view of the entire range of RAM, it's all just a way of jury-rigging more memory onto a system not designed for it.

True, it may not be as hard to use as segments, but it's not as easy to use as 100% flat addressing either. And the 4gb limit is still a problem for some RDBMS systems.

Order of magnitude.

[Inoshiro]

Let's say I have a 16,000 by 16,000 image frame. I have 60 of them per second, and have 2 hours of film. That works out to 100.5 terabytes of data. 2^64 happens to let me store and address about 166,937 of these superduperHD movies.

Now, I don't know about you, but I only own a couple-hundred movies, and I only own a couple-hundred games. Even if they were the mega, mega high res I mention above, I'd still not use up more than a miniscule fraction of what I had available. That's why I think it'll last at least a century.

Re:Order of magnitude.

[kilonad]

Alright, I suppose I can see where you're coming from now. I'd love to know how much space holographic video would take up, but I doubt that will be coming out in the next 50 years, especially not at the consumer level, so there goes that theory. So you're right that it'll probably last at least a century. You are, however, incorrect on your math. You computed the size of an 8-bit grayscale video. The real size of the 24-bit uncompressed video would be around 301.75TB. 16,000x16,000 pixels/frame * 3 bytes/pixel * 60 frames/second * 7200 seconds = 301.75TB. And that doesn't include audio, although even 6 channel 96KHz 24-bit audio would only cost you 10.9GB for two hours. Of course, this is all in the world of the fanciful, where compression doesn't exist. We're likely to see fractal compression in common use in 50 years, which will allow for compression ratios of at least 30:1 at virtually perfect quality, so even those two hours of mind-boggling video would take up no more than 10TB. I'd love a drive that could write the 2.2GB/sec that uncompressed 244MPixel video would require. You'd surpass the old AVI file-size limitation in under a second. :)

TACO AND HEMOS ARE GHEY

[Anonymous Coward]

Get it in ya!

Re: Skip to the last page for the most interesting

[Cyno]

Interesting. I was considerring upgrading one of my systems, but I may just hold out and wait for that PPC 970 chip. Bet it'll render MPEG-4 as good as any AMD, and I mostly like IBM these days.

The 64-bit desktop

[Jugalator]

"He also thinks there will be a market for desktop 64-bit systems."

I'll only say this...

http://www.theinquirer.net/?article=6501

Re:+5 Blatant Karma Whore

[ameoba]

I was hoping. If I could get some karma out of my CS degree, it'd be more than I've been able to get from it yet, considering the dismal state of the economy. Now is not the time to be a fresh graduate.

Killer applications

[pacc]

"early adopters, people like the gaming folks. Those people are really dumping tons of data into broadband"
'nuff said; AMD is really counting on the DVD-iso scene demaning higher performance..

Kudos for the direct link to last page, that should be the slashdot standard for any story and hardware review!

NOT Moore's law...

[Anonymous Coward]

People who misunderstand "Moore's Law" make me angry.
First of all, it doesn't state that every single number ever associated with a computer will double every 18 months. It pisses me off that people try to apply it to every single thing. "Windows 2000 just came out? I guess that means Windows 4000 is coming out in 18 months! Moore's Law says so!" Moore's Law has to do with transistor counts in CPUs, and so is completely and entirely irrelevant.
Secondly, it's not a law, not nearly, and it's only called a law to be ironic. It's an observation of historic trends, and so could be used to fairly reliably make short term predictions, but using it to predict something out to 2050 is psycho. IIRC, Moore's Law started out with a doubling period of 2-3 years, and crept down to 18 months. What makes you think that's a constant? What'll happen when more countries get involved with computers?

interesting

[Dave_bsr]

I know newegg's a pain for links...sorry.

i've never had problems with newegg or that chipset...2/2 times. *shrug*.

But I HAVE had troubles with PCChips, two boards, both awful, with onboard video etc, no support, and general crappiness. My Mom's PC that i didn't get to suggest any change about, and a friend of mine who's main questions was: how cheap is it for what sounds good? ARGH! PC chips is a whore.

So... ECS is PC-chips? huh. something new...

Liteon: I've never had problems. Maybe I'm lucky. never know.

Back to homework.

Re:interesting

[shepd]

>So... ECS is PC-chips? huh. something new...

I think so. Imagine this board with this board's colours. It's a spitting image! :-)

Actually, the K7S5A has turned out to be far more durable than any other PC Chips board I've ever met. I would actually say it's a fair board, but I just can't knowing PC Chips past.

So don't worry too much, just count yourself lucky! :-)

More, bigger registers for 64 bit, not 32 bit

[mazor]

The additional registers and the wider register sizes do nothing for 32 bit applications running in the AMD's 32 bit x86 mode. A wider data bus will probably benefit 32 bit apps, but other than that, to get the *full* benefit of an AMD-64, you'll need applications recompiled for the AMD-64 bit instruction set.

That's the same as for the Intel Itanium. The AMD-64 architecture has two HUGE differences from the Intel Itanium setup:

1) the AMD-64 instruction set is a rather minor modification/rearrangement of the traditional x86 opcodes. That means existing x86 optimizing compilers should be able to produce AMD-64 executables in a matter of days, compared to the years it's taking to get decent Itanium codegen compilers.

2) The AMD-64 runs 32 bit apps in hardware 32 bit mode concurrently with 64 bit apps and OS running in hardware 64 bit mode. The Itanium doesn't get anywhere close to that - At best, 32 bit x86 code is emulated (slowly!) in software microcode on the Itanium chip.

If AMD can produce 64 bit chips that run 32 bit software as fast or faster than current 32 bit processors, at a price point near the high-end 32 bit processor prices, AMD will mop the floor with Itanium...

-mazor

Reasons to ditch x86?

[Decimal]

What is so bad about x86 anyhow? Is it that its support for smaller sections of registers (8 bit AL register that merged with 8 bit AH register to become the 16 bit AX register, then EAX, RAX, etc) made for more hardware that slowed it down? That's apparently not the case, with the speed war between AMD and Intel pushing ~3 GHz. So why does x86 have such a bad rap? Because it is little-endian?

Re:Reasons to ditch x86?

[GooberToo]

It's generally agreed that it's the limited number of available registers and how the available registers may be used (limited number of general purpose). The limited number of registers is one of the reasons why it's generally considered slightly harder to heavily optimize for x86 than many other CPUs. IMOHO, any complaints beyond what's being described above is simple CPU-bias.

Cache pressure.

[TheLink]

Actually cache pressure may not be that bad. Correct me if I'm wrong.

For OS caches, 64 bit is neutral, what you lose in bigger data, you gain in being able to have more RAM.

With respect to CPU caches, lots of data sets don't fit anyway. Unless you're talking about 8MB caches - very expensive .

In contrast, my guess is the size of loops in program code haven't increased, in fact they probably have gone down due to caching (Don't want to unroll a loop till it falls out of the various caches).

Instantaneous = ~0.05s

[Anonymous Coward]

You won't be bothered with a computer that uses more than 0.05s to respond. Consider how bothered you are with the 24 frames/second that films use. Are you bothered with the delays when you watch a movie?

So 0.05s is the target.

Browsing the _Intranet_

[TheLink]

Not just high-end games or 3D modelling.

Sure I agree that it's not worth upgrading from 2.4GHz to 3GHz, but can't tell the diff between 866MHz and 2.4GHz???

You do notice faster processors when you use scripting languages :). Or databases.

Believe me you'll feel the difference when _serving_ the web vs surfing the web.

As for web browing? Faster machines on both client and server make the _Intranet_ browsing experience VERY different.

When you match servers which can serve up dynamic custom pages at video refresh rates (70 hits/sec)with clients that can render it at that those rates. Pages just snap up in front of you. Very very nice. Your application stops getting in the way of people and users stop noticing your application even subconciously.

Users who really don't notice the difference even when switching to slower machines are the same users who don't notice what the dialog box said before they clicked OK, and will then tell you they didn't change or do anything.

Translation

[TheLink]

Why so many words?

Dave _bsr's Law: They'll never be fast enough.

Computer, count to from 0 to infinity and finish it now.

Not useful? OK how about this:

Computer, do a statistical analysis of prime numbers vs normal numbers from 0 to infinity and tell me if you find anything interesting based on my previous interest in this subject, NOW.
--

Heck even the speed of light isn't fast enough. Ask any transcontinental quake player.

What an abominable article

[Keith_Beef]

How cheap can you get? You take an interview between some second-rate reporters (pompously described as being "Executive Editor/News" and "Editor in Chief") and some marketing exec who cannot talk coherently, and just transcribe the conversation. Leaving aside the travel costs, you've got an article for the cost of twenty minutes' worth of typing secretary.

Of course, Linux is already way ahead and on top of Opteron. Their far ahead of anybody at the moment ...

The monkey of an editor didn't even proof-read!

I'm enthusiastically waiting for Hammer (I'm waiting as fast as possibly I can)... but this kind of article makes me glad I don't shell out any schekels for eWeek.

And 640k is all the memory we'll ever need!

[Monofilament]

hmm don't we remember someone saying that once upon a time.....

Re:And 640k is all the memory we'll ever need!

[ppetrakis]

Sir,

If you think making this move is as easy as adding more ram to your system. You'll find out how wrong you are real soon in the most painfull of ways. When the sourcecode for project X you just compiled doesnt work or even gets that far! You obviously didn't understand my post. I was talking about the big picture, not don't buy AMD. But hey just like kids who don't listen to their parents, you'll find out the hard way :).

Peter

Re:And 640k is all the memory we'll ever need!

[Monofilament]

Well yes I do understand the logistics of moving to 64bit architecture. I know its a pain in the ass, but I think the long term benefits of architecture change, will outway the pain it causes now. Whole new generations of coders will grow up never careing that there was a 32bit architecture before. 1. It was a joke..ligthen up. 2. I'd say its short sited to believe just because moving to 64bit is a hassle, its not a benefit. Innovation has its price, and many times it involves a lot of hard work, but thats no reason to vegetate in the current medium just because that what you use now, and you can't see any good use right now. But I also saw in the article the fact that AMD Says it will be reverse compatable with 32 bit legacy software.. maybe I didn't understand the original article.. but thats what I got out of it. So it wouldn't even be that painful. Now yes to benefit MORE from 64bit you'd have to change your code or make up a whole new program.. but hey we make new versions of software all the time anyway..

Re:And 640k is all the memory we'll ever need!

[ppetrakis]

The AMD is reverse compatible with 32bit and the two can coexsit granting the host OS has both supporting 32 bit and 64 bit libraries, Alot like Solaris. What makes Hammer's introduction so critical is the potential to upset the entire balance of the whole PC market. I'm not telling you 'not' to do it. I telling you what the future may hold. Take for example openoffice. Now Sun itself won't dare port that to 64 bits even for it's own OS because it's just too damn big. They're using the same "tricks" we use in Alpha land to compile 32bit programas to use 64 bit addressing. There's a HUGE performance hit incured when this is run their true 64 bit chips. Their older ones, alot like the hammer can turn that extension on and off without penalty.

Now let's take another company for an example that will probably be pushed by it's userbase to port to 64, Adobe. These are the same guys that said no to recompiling photoshop on NT/Alpha. Remember that NT on Alpha is a 32 bit OS, there's no 64 bit isms to deal with. Now, When you have to invest to port an application like photoshop to 64 bits. The work is so huge that you can be open to all sorts of possibilities including adding/removing platforms. When the move is that big, why not consider it???

Major application vendors could tell MS to take a hike and move to Linux or whatever does the best job at supporting a 64 bit architecture. This also puts those same vendors in some extreme risk. What happens if the competition gets their port done before yours? What could it cost the company? Fear of new upstarts with 'clean' codebases who can get their product to market faster, with more "feathers in their hat", and no legacy apps and plugins to support. It's huge.

64 bit isn't next gen or even future gen technology, It's "then gen". 10 years now Alpha has been out there. There are some companies out there who already have 64 bit ports done or on the way with no public knowledge. Ready to play that trump card on the competition.

If you own a hammer and their are two apps you're considering to buy. One is 64 bit and one isn't, Which one is going to get more consideration from the start? That's the big deal and that's why I'm so serious about it. When you're not bound by platforms anymore it's a function of "what works best?". I could be Linux/Hammer, PPC64/MAC, Itanium/HPUX, or even AlphaLinux.

Peter

Re:And 640k is all the memory we'll ever need!

[Monofilament]

He he ok you got me I give up. I think we've been talking past each other but I figured I'd have some fun.. and just say I wanted to get the last post in this string of replies.. though you'll probably reply anyway.. and I'll keeping on replying back until they stop the thread..

who knows what the future holds?

the game begins..

Actually ....

[fleppir]

you can have 7 of those without breaking the limit :)

Re:What desktop users want to know..(OFF TOPIC)

[fredrik70]

HEY!
Don't talk bad about vinyl! All decent tunes are still released on 12 inch first and sometimes only!

oh, maybe I should get a live part form playing records....

"It's the end user, stupid"

[melonman]

technology is now not an issue. You can do almost anything you want to with technology. Can we now make it more useful? Can we make it more practical?

AMD will sell a certain number of chips because 64-bit is a sexy number, just like the obsession with clock speeds, just like people buy PCs with 32Mb of video ram to do WP.

But the quote above seems to me to encapsulate the challenge facing the IT industry in general. What qualititive difference does the latest processor - or the latest kernel for that matter - make to the end user? What real-world job will it let me do that I can't do now? In terms of servers, it is easy to answer the question. In terms of desktop machines, I'm not so sure.

parent links to goatse

[Anonymous Coward]

parent is the most fucked up browser trapmine i have ever witnessed.

don't try this in your computerlab or at work. or even better: send it to your friends at work.

64bit registers

[TwistedSpring]

The advantage here IS, in fact, in speed. The reason is with 64bit registers you can store more information in a register: twice as much as with a 32bit CPU. This is an incredible advantage. If you're writing software in assembler, or compiling software with a compiler that is 64-bit aware, you have twice as much space on the chip to play with. Registers are so much faster than RAM that to compare them is simply meaningless. With something that has been sensibly coded and makes use of this extra space on-chip, the speed difference will be phenomenal. You can now store much larger, higher precision numbers in the CPU registers. 64-bit has a similar effect to doubling the number of registers in a 32bit processor. You can therefore keep more of your mathematics on the chip, and you dont have to swap data between the chip and external RAM as often. This is a colossal benefit.

Some people have mentioned that 64bit allows you simply to address more RAM. Registers arent used simply for addressing RAM, and currently the consumer market doesnt typically have 4gb of RAM to play with.

Before 64-bit you would have had to use external RAM to store an extremely high precision number, or two 32-bit precision numbers, and external RAM is unbelievably slow, even L1 and L2 caches are unbelievably slow compared to register speeds. Keeping twice as storage space on-chip means that (if it's exploited correctly) the speed difference will be huge.

Of course, with current software development ideals we will have to see some extremely intelligent compilers before this is exploited to it's full potential. And it will be a long while before 64-bit takes off, despite what AMD say. They can release the chip, but if there's no software that exploits 64-bit technology (and software developers will fear releasing software that only 64-bit CPU owners can run) then the chip will be useless, and I think consumers will realise this. Ok maybe they won't. They'll see 64 bit and think "wow! my 32bit programs will run so much faster!" but what the hell, if it increases AMD's market share then I am all for it.

Re: Can anybody compare and contrast the two ...

[Ninja Programmer]

Short answer:
Itanium = IBM PS/2. Hammer = PC AT.

Long answer:
Itanium is a VLIW architecture with a new 64 bit instruction set, that performs in-order execution (like an original Pentium, or a Sun UltraSparc.) It has very high floating point performance, because of its dual, fully pipelined FMAC units. The architecture is interesting in that it uses a wide variety of mechanisms (predicates, conditional loads/stores) to avoid branches without using speculation.

Because of its new instruction set, its not compatible with any prior software. *Everything* must be ported from scratch, and there is no prior install base from which to leverage. (Actually that's not strictly true -- it has an x86 compatibility mode that is exceedingly slow; its like a quarter to a third of the performance of leading x86s.)

Compilers in particular are a really different kind of beast on the IA-64. All the "out of order execution" techniques that are in most other modern CPUs have to be somehow implemented in the compiler. The performance of everything depends on these compilers, and so far only Intel and HP's compiler have measured up to snuff. SGI's, Microsoft's and gcc have been ported, but apparently aren't very credible. It appears as though creating good compilers for IA-64 is a truly monumental task.

Microsoft has gone on record as saying that they have no intention of porting Office to Itanium -- its not that kind of platform. Itanium is seen as a server only player. While one is tempted to say "oh but that will come to the desktop eventually", unless Microsoft (or Apple or Sun -- but those are even less likely) is willing to port a significant number of applications to it, that's just not going to happen. Microsoft has ported Windows to IA-64, but has pretty much stopped development of IA-64 at that point.

Itanium has been available for over a year now, and has had a very cold reception in the marketplace. Nobody wants them.

Hammer (Opteron) is just a 64 bit extension of the x86 architecture. If you are an assembly, or system level programmer take a look at this:

http://murl.microsoft.com/LectureDetails.asp?690

Its quite detailed. The key thing to note is that the x86-64 instructions are really just extensions of the IA-32 instructions. That means that all the compilers and tools work almost identically to how they worked for IA-32. The key features that have been added are: 64 bit registers, a 64 bit address space, and twice as many integer and SSE registers.

x86-64 remains backward compatible with ordinary x86 in two ways. If the chip boots as normal with a 32-bit OS, that is unaware of its 64-bit mode, than just like 16-bit DOS before it, it functions like a normal old 32-bit Athlon. It also contains two other interesting modes: 1) Long Mode (this is how 64-bit are enabled) and 2) Compatibility Mode.

Long Mode enables the new instructions, registers, address modes, etc. Compatibility mode allows it to execute the old 32-bit software in a virtual environment, much like the v86 mode used for DOS Boxes under Windows.

It is known that Microsoft has ported Windows to Hammer, and apparently Office and Internet Explorer are up and running on it. But don't be fooled, Microsoft didn't waste their time *porting* Office or Explorer to x86-64. They simply run in compatibility mode under this new Hammer enabled 64 bit Windows. That's the whole beauty of their scheme -- 32 bit and 64 bit applications can be running at the same time, sharing the same OS resources.

Now because x86-64 is so similar to IA-32, it allowed AMD to implement both long mode, and compatibility modes in the same pipelines, with optimal ALU usage, using the same instruction translation super-structures, the same rename registers, etc, etc. I.e., any effort AMD spends in speeding up the critical paths of 64 bit operations, translated back directly to 32 bit operations and vice versa. There is no compromise -- it will be a very fast 32 bit processor, as well as a very fast 64 bit processor.

While Hammer will not be quite at the same performance as IA-64 for floating point, it won't be that far behind, and it will almost surely cream it on integer performance. This much has been revealed by AMD/Intel's Spec CPU claims/disclosures thus far.

As far as selling Hammer? Its just a natural follow-on to Athlon. It won't cost more (it actually has a smaller die size, so it actually costs AMD *less* to make Opterons than Athlons) and it will be just as compatible with your software (Linux, Windows, or whatever.) At the very least, people *will* adopt them for the same reason they adopted the Athlon. So 64-bit ready desktops are going to ship middle of next year, and the customers *will* be lining up to get them (regardless of whether or not they use the 64 bit features, or even get the 64-bit Windows for it.)

Microsoft has gone on record as saying they endorse AMD's approach. And of course they don't need to commit one way or another to porting any applications, since they will all run at top speed on Hammer regardless. While there is nothing more overtly committed by Microsoft, reading between the lines will lead you to realize the Microsoft probably intends to port its DB and IIS to Hammer.

Linux, of course, has been ported to both, but I don't know if they have enabled the compatibility mode for Hammer.

IP-relative addressing

[Anonymous Coward]

The x86-64 also has decent IP-relative addressing. With a new standard for shared libraries, this could make shared libraries MUCH MUCH faster, rather than the heinous kludge that is ELF.

good solid business advice, jon.

[twitter]

Alpha has been in production for 10 years. How come it's not cheap? What? The lowest common dominator company don't like it? That's what I thought.

When you think it's OK to screw people for buisness reasons, that's what you will get.

Re:good solid business advice, jon.

[Anonymous Coward]

you misinterpreted what i said, you fscktard.

Re:good solid business advice, jon.

[JKR]

Alpha is minority because post-VAX DEC couldn't market their way out of a paper bag. Look at the DEC Rainbow (their answer to the PC-XT, and technologically a much better machine). They failed to compete with IBM & Compaq (the clone factory), and that pretty much killed them.

The Alpha legacy lives on as one of the best performance / power ratio processors, the StrongARM (and XScale) both of which use techniques developed for Alpha.

I don't think it's OK to screw people in the name of business, but I don't think you should bend over and take it either...

Jon.

Needed for Memory-Mapped files!

[DoronRajwan]

Recently, I developed a tool for Forward Error Correction codes (FEC). When encoding and decoding packets, I use memory-mapped files, since I need random access and virtual memory optimizations, as done by modern OSs.

Working with up-to 1GB was simple. I memory-mapped the entire file, and access it. In order to break the 2GB barrier (the user-space virtual address space size), I had to implement a sort of "cache" that remembers the last areas I accessed, and map only those areas into memory. I had to map and un-map areas when switching from one area of the file to another. Instead of a clean and efficient code I now have ugly, complex, and bug-prone code.

I think having 64-bit virtual address space is critical for today's applications.

Just a note: The Windows virtual memory manager is horrible. Developed for Linux 2.4, our product worked great. Random file access decoding, of 1GB file, on 100MHz Pentium-I based set-top-box with 64MB RAM, and slow hard-drive -- at ~10mbps. When ported to Windows 2000/XP, a 700MHz Pentium-3 with 256MB RAM and modern hard-drive could not even get close to the set-top-box. I am talking about a factor of x20 in performance!

The Windows version didn't commit pages to disk -- it simply ate more and more memory, until the computer became unworkable. Alternatively, when removing the memory-mapped from a part of the file, Windows did a sync() on that part, caused a delay, and an x10 performance penalty. A lot of optimizations were needed to overcome these problems. Still, the Linux version is x3 faster. (I didn't test kernel 2.5 yet).

One bit per year

[IPFreely]

2^64 is certainly enough for at least a hundred years

In the late 70's, we hit the limit of what 8-bit computing could do. 16-Bit processors were produced.
8 Years later 16 bit processors were hitting their limit. In '86 Intel produced a 32 bit processor.
Now, 16 years later, we will jump from 32 to 64 bit processors.

That should last us for another 32 years, then we'll hit the limit and really have to go with 128 bit. A Quick calc of the needed bits would be the year - 1970. Sure, we may not need 64 whole bits right now. But we do need more than 32. so the next obvious step is to just bump it up to the next power of two and wait it out.
(This is desktop microprocessors of course. High end and mainframe processors have advanced at earlier dates, but about the same intervals.)

Thank You

[Erioll]

Thanks. This is exactly the kind of answer I was looking for. I'm glad that the problem has been tackled on other architectures at least. I don't know anything about sparc, and didn't know that they had already done something similar to what AMD is doing (different instruction set, but same general idea).

Thanks for answering my question.

Erioll

Re:Heat. tsarkon reports.

[Jim Norton]

I see.... the P3 never had any heat issues beyond 1 GHz whatsoever, even though they recalled the 1.13 GHz due to exactly that reason (this was BEFORE the Tualatin/0.13 micron shrink)

Seriously...

Now the biggest problem with your filth is that Hammer specs aren't out. system integrators need thermal information to make designs. And when Dell and friends get the thermal data, along with everyone else, then you can talk about what kind of heat output AMD-64 will have.

No shit. Didn't you even read my post? I said there was no definitive answer to his question... oh, never mind. We've been down this road before and all i'll get is more babble and garbage posts laden with personal insults i've already heard from you a million times before which were taken right out of some Elementary school playground.

Grow up. Seriously. It's been a month. A fucking MONTH and you still haven't gotten the point yet. I'M NOT LISTENING TO YOU. Get a life! Do something productive instead of wasting time spewing crap and typing shit nobody is interested in reading anyway.

Re:Heat. tsarkon reports. right on jim!

[Anonymous Coward]

i mean, i was reading your history and stuff, and i was like looking at this guy you know and he was like talking all this smack and you were like being innocuous or friendly, but he was like being acidic and stuff, and so like you've been dilligent and stuff and you tried to ignore him and he just keeps its up and this guy is well, you know, persistent and stuff and you were like telling him like, "go away troll dude," and he was like saying the 7 bad tv words and like you said "like dude, yo, like lay off man," and he was like doing more of the 7 word stuff, and you were like, "dude, im not talking," and he like still keeps talking you know and like is its seriously messed up. well good luck fighting off the troll. I'm the founder of the Anti-Troll Society. In this discourse, I will tell you what made me form such an organization and how I plan to use it to stop defending the homophobic status quo and, instead, implement a bold, new agenda for change. Although not without overlap and simplification, I plan to identify three primary positions on Troll's conclusions. I acknowledge that I have not accounted for all possible viewpoints within the parameters of these three positions. Nevertheless, it's really not bloody-mindedness that compels me to enable patriots to use their freedoms to save their freedoms. It's my sense of responsibility to you, the reader.

His idiotic claim that freedom must be abolished in order for people to be more secure and comfortable is just that, an idiotic claim. Look at what's happened since Troll first ordered his stooges to kill the goose bearing the golden egg: Views once considered insidious are now considered ordinary. Views once considered inarticulate are now considered perfectly normal. And the most negligent of Troll's views are now seen as gospel by legions of jaundiced slobs. If everyone does his own, small part, together we can tell him where he can stick it.

If I weren't so forgiving, I'd have to say that even those few who benefit from his allegations fail to recognize their current manifestation as a mawkish form of cynicism. That concept can be extended, mutatis mutandis, to the way that if I withheld my feelings on this matter, I'd be no less hate-filled than Troll. Imagine a world in which he could wipe out delicate ecosystems whenever he felt like it. In the course of my work, I regularly come in contact with the worst classes of contemptuous freebooters there are, and most of them also feel that we must create greater public understanding of the damage caused by Troll's perceptions. As mentioned above, however, that is not enough. It is necessary to do more. It is necessary to provide some balance to Troll's one-sided slurs. I just want to sound the bugle of liberty. That's why I propose, argue, cajole, plead, wheedle, and joke about ways to make this world a better place in which to live. Next time, Troll, you may want to check your facts correctly. Faced by such despicable perfidy and the frustration of not being able to respond to the same audiences as Troll has had, I must indubitably reinforce what is best in people. And that's why I say to you: Have courage. Be honest. And deal with Troll appropriately. That's the patriotic thing to do, and that's the right thing to do.

* J i m N o r t o n * F u c k s * A s s ! F a g *
Jcccccccccccccccccccccccccccccccccccccccccccccc cJ
ic/ccccc\ccccccccccccc\cccccccccccc/cccc\ccccc cci
m|ccccccc|ccccccccccccc\cccccccccc|cccccc|ccc cccm
*|ccccccc`.ccccccccccccc|ccccccccc|ccccccc:c cccc*
N`cccccccc|ccccccccccccc|cccccccc\|ccccccc| cccccN
oc\ccccccc|c/ccccccc/cc\\\ccc--__c\\cccccc c:cccco
rcc\cccccc\/ccc_--~~cccccccccc~--__|c\ccc cc|ccccr
tccc\cccccc\_-~cccccccccccccccccccc~-_\c ccc|cccct
occcc\_ccccc\cccccccc_.--------.______\ |ccc|cccco
ncccccc\ccccc\______//c_c___c_c(_(__>c c\ccc|ccc cn
*ccccccc\ccc.ccCc___)cc______c(_(____>cc|cc/cc c c*
Fccccccc/\c|cccCc____)/Norton\c(_____>cc|_/ccc c cB
ucccccc/c/\|cccC_____)Fucks *|cc(___>ccc/cc\ccccl
cccccc|ccc(ccc_C_____)\_ASS __/cc//c_/c/ccccc\ccco
kccccc|cccc\cc|__ccc\\____ _____//c(__/ccccccc|ccw
scccc|c\cccc\____)ccc`--- -ccc--'ccccccccccccc|ccs
*cccc|cc\_cccccccccc___\ ccccccc/_cccccccccc_/c|c*
Accc|cccccccccccccc/ccc c|ccccc|cc\cccccccccccc|cG
sccc|ccccccccccccc|ccc c/ccccccc\cc\ccccccccccc|co
sccc|cccccccccc/c/ccc c|ccccccccc|cc\ccccccccccc|a
*ccc|ccccccccc/c/ccc ccc\__/\___/cccc|cccccccccc|t
ccc|ccccccccccc/ccc ccccc|cccc|ccccccc|ccccccccc|s
ccc|cccccccccc|ccc cccccc|cccc|ccccccc|ccccccccc|*
* J i m * N o r t o n * SUCKS Fucking Choad * Fag

Re:Heat. tsarkon reports.

[Anonymous Coward]

I have polled others; you elicit much the same response that I have for many others. We are not babies, playgrounders, or fat sexless E.Q. rpgers or ass philanderers such as yourself.

Your stupid snarky ad hoc anecdotal crap is just lame. You say stupid shit all the time. And now I dump and piss on you for cathartic purposes, and I get paid a great salary to do it [and I still get good performance reviews ;p], which makes it more fun, then I get others to read me lambaste you and we laugh at your expense. All I ask is that you keep posting, its such great fodder. In fact, I encourage more retaliatory responses. They are priceless, and reveal your stupidity, weakness and frustration.

Re:Heat. tsarkon reports.

[Anonymous Coward]

In fact, Jim, the stepping of that CPU you mention is:
S-SPEC: SL4HH core stepping cC0 CPUID:0686h

What does this have to do with the fucking hammer? the barton? the athlon?

I don't see the parallel - at all! In fact, Barton is similar to the P3 Tualatin, you insinuate they ran out of rope and have nowhere to go. How to you explain the tAx steppings if the P3 was so out of gas? Huh?

You are such a gamer fan boi. Fucking gaywad.

Re: Skip to the last page for the most interesting

[Elwood P Dowd]

Craig Mundie (perhaps a different MS PR Rep) said in an interview that he felt the GPL was unamerican, and suggested it might be something that congress should "look into". Dunno if MS has actually done anything about it, but comments like that plus campaign contributions might be all that's feasible. Don't want to go googling for links right now, sorry...

Intel has done no such thing, of course. I don't know what the hell the troll meant by saying that they were "backing" orgs trying to destroy OSS.

Last Post!

[alpg]

I forgot to mention an important fact in the 1.3.67 announcement. In order to
get a fully working kernel, you have to follow the steps below:
- Walk around your computer widdershins 3 times, chanting "Linus is
overworked, and he makes lousy patches, but we love him anyway". Get
your spuouse to do this too for extra effect. Children are optional.
- Apply the patch included in this mail
- Call your system "Super-67", and don't forget to unapply the patch
before you later applying the official 1.3.68 patch.
- reboot
-- Linus Torvalds, announcing another kernel patch

- this post brought to you by the Automated Last Post Generator...