First Round of AMD Athlon 64 Reviews In

Memory mapped disk? by Thinkit3 · 2003-09-27 06:02 · Score: 1

Isn't this a major benefit of more bits? You can play with the concept of mapping disk like memory, since you have plenty of address space for any size disk. The Athlon 64 might be a kludge, but introducing more bits is a good thing.

--
-Libertarian secular transhumanist

Re:Memory mapped disk? by Kenja · 2003-09-27 06:05 · Score: 1

Given that there is no way to fit more then 8GB on an ATX MB and that requires spending 700$ for each of the eight 2GB DIMMs I'd say that the extra "bitness" is overhyped. You will not see consumer systems with more then 4GB any time soon.

--

"Have you ever thought about just turning off the TV, sitting down with your kids, and hitting them?"
Re:Memory mapped disk? by Anonymous Coward · 2003-09-27 06:14 · Score: 3, Informative

If you have lot of RAM or so programs will tend to exhaust the 32-bit address space before actually consuming all the physical memory. One example is that creating a thread might represent the reservation of 64KB of address space for its stack, while only 8KB of physical memory is actually committed.
A 64-bit address space is probably a good thing once a program is allocating 2GB or more of address space.
Re:Memory mapped disk? by DetrimentalFiend · 2003-09-27 06:25 · Score: 1

That's true, but new form factors are on the horizion and extra bits now means that when form factors do more memory, you won't have to patch the software and redesign CPU's again.
Re:Memory mapped disk? by ianezz · 2003-09-27 07:02 · Score: 4, Interesting

You will not see consumer systems with more then 4GB any time soon
The point about mmap(2) is to let the system (the VM subsystem of the kernel) manage the caching for the userland processes using it, avoiding extra copies to/from buffers in userland and eliminating in several cases the need for custom caching code (processes don't have to worry about data being available in RAM: the kernel automatically takes care of that when needed).
You don't need gobs of memory to do this, but in order to work on large amount of data you need a large address space, which is what 64 bit architectures provide. Of course, the more physical memory you have, the less the kernel has to swap pages in and out, but the main point is not that.
A little example to clarify: in order to keep things simple (instead of needing two intermixed caching systems, one for the VM and one for disk accesses), the Hurd just mmaps the whole partition. This means that the maximum size of a partition has an upper limit given by the size of the addressing space, which is 4GB on 32bit architectures (actually less, since in that address space you have to keep also the code that uses the mmapped data, so it's more like 2GB/3GB). A 64bit architecture comes very handy here, given the current size of hard disks.
Re:Memory mapped disk? by ewoodlief · 2003-09-27 09:36 · Score: 1, Insightful

The Athlon 64 is revolutionary in that not only is it way ahead of its time, but it will also accelerate all other product industries such as motherboards, memory, data storage, software, etc. Yes, now, prices of components required in order to leap to this level of technology that will fully utilize its capabilities is expensive. Over time, prices will fall as those products will become the next standard.

AMD is able to release a cutting-edge processor while at the same time get away with telling the other markets to get-with-the-program because of the most ingenius aspect of the Athlon 64: it runs in both the 32-bit and 64-bit domains. Therefore, in the interim, we can upgrade simply to take advantage of faster bus speeds and multimedia instructions on an already proven chip, namely the Athlon. And down the line, with is closer than you think, there will be a chip and a company, with experience, welcoming us into a new era.
Re:Memory mapped disk? by chiph · 2003-09-27 10:50 · Score: 1

The AS/400 has been doing this for years. It's (technically) a 65-bit system, as it reserves one bit to differentiate between user-mode and OS-mode memory. All hard drives (DASD in IBM-speak) are mapped into this memory address space, and the memory caching subsystem is responsible for making real the address you requested. Maybe it's in memory, maybe it's on disk, but to you, it's always going to be 0x0034a4b37cff0801.

Chip H.
Re:Memory mapped disk? by mdowling · 2003-09-27 14:25 · Score: 1

"You'll never need more than 640k of memory..." -Bill Gates
Re:Memory mapped disk? by DrMrLordX · 2003-09-27 15:19 · Score: 1

How many of these CPUs have had any major impact on the desktop market in the US or abroad? I know that EV-6 was originally used for Alphas and later turned up on Athlon platforms, so score one for the Alpha.

The Athlon 64, while nothing terribly new or novel, is causing a major upheaval in the desktop market. It may, and likely will, lead to the mass-adoption of 64-bit OSes and software on PCs everywhere. If people won't be running the software on an Athlon 64, they'll be running it on Tejas. You don't honestly think Intel would have put 64-bit extensions on Tejas without the Athlon 64 in the marketplace, do you?
Re:Memory mapped disk? by cout · 2003-09-27 15:35 · Score: 1

The point about mmap(2) is to let the system (the VM subsystem of the kernel) manage the caching for the userland processes using it, avoiding extra copies to/from buffers in userland and eliminating in several cases the need for custom caching code (processes don't have to worry about data being available in RAM: the kernel automatically takes care of that when needed).
Most applications, however, do not take full advantage of mmap. Most applications don't even need to take advantage of mmap. Those that do use mmap for something other than memory allocation (malloc uses mmap under the hood) or loading shared libraries don't mmap large files. It's just as simple to use a database for most applications, plus with a database you get all sorts of querying abilities without the extra development time it would take to do the same with an mmap-based solution. And databases don't get dumped with the rest of the core when an application segfaults (on Linux, mmap'd memory gets dumped to disk with core, and afaict there's no way to disable this).
I seriously doubt we'll be seeing applications any time soon taking advantage 64-bits of address space by using mmap().
Re:Memory mapped disk? by be-fan · 2003-09-27 17:40 · Score: 1

The practical memory limited for a consumer 32-bit machine is not 4GB but 2GB. The OSs needs some of the virtual address space (you can cut it down, but that decreases peformance because you have to jump through hoops to map the whole buffer cache) and hardware needs some as well. All told, up to a gig of your virtual address space is gone before you even talk about memory. Since 3GB is an odd amount of memory for a machine, 2GB is the limit you usually see in motherboards. 2GB is a pretty close limit. Soon as my new SO-DIMMS get here, my *laptop* will have a gig of memory. My next system will probably be 2 gig, and that leaves me no more room to expand. So the 32-bit limit is getting close.

--
A deep unwavering belief is a sure sign you're missing something...
Re:Memory mapped disk? by Emil+Brink · 2003-09-27 22:00 · Score: 1

Maybe that's so. But at least you could cut the cost in half by realizing that you can build 8 GB from only four 2 GB modules! Amazing, huh? *Ducks*. ;^)

--
main(O){10<putchar(4^--O?77-(15&5128 >>4*O):10)&&main(2+O);}

Hmm.. by darkov · 2003-09-27 06:03 · Score: 3, Funny

Is it possible to have a first round of reviews after an earlier story about Athlon 64 reviews?

fp

--
Reliable, Great Value Hosting: $7.95/mo 2.4G/120G

Re:Hmm.. by bj8rn · 2003-09-27 06:10 · Score: 2

Yes it is:
1) The earlier story was round 0, or
2) the earlier story was the pre-match bragging part, or
3) some other reason (not profit, though)

--
Hell is not other people; it is yourself. - Ludwig Wittgenstein

Wooah! by Sir+Haxalot · 2003-09-27 06:04 · Score: 1, Funny

Remember bigger numbers =! better.
Think N64 and Playstation.

--
I have over 70 freaks, do you?

Re:Wooah! by sirmikester · 2003-09-27 06:26 · Score: 1

While I agree with you in concept, your example has a flaw.. the N64 had far better performance than the playstation. In the case of performance bigger numbers == better.

--
In linux libertas
Re:Wooah! by Anonymous Coward · 2003-09-27 06:50 · Score: 2, Funny

Correction =! != !=
Re:Wooah! by Sj0 · 2003-09-27 11:25 · Score: 1

Woah! Now if you could only run that emulator with the fancy filtering on the Playstation, your example would actually be relevant!

If you don't have the reading comprehension of a 12 year old, why bother posting? In fact, why bother reading in the first place?

--
It's been a long time.
Re:Wooah! by 0000+0111 · 2003-09-27 13:25 · Score: 1

Correction =! != !=

Unless you're Yoda

64bit.. Schmobit... by deunan_k · 2003-09-27 06:04 · Score: 2, Funny

Enough! Leave me alone with my abacus...

Um, anyone knows how many bits can an abacus counts to?

--
Will sys-admin for food

Hang on.. by madprof · 2003-09-27 06:06 · Score: 3, Insightful

Give it a bit more time. The motherboards can still be tweaked to get more performance out of the system. I think we should wait a bit before really making decisions, and get in a whole load more real world testing too.
Benchmarks are not always entirely, although often can be, illuminating.

Re:Hang on.. by Gldm · 2003-09-27 06:28 · Score: 1

The FX will also probably perform a bit better once it's freed from having to use registered memory. Wait for the 939pin version.

--
Introducing the new Occam Fusion! Now with sqrt(-1) fewer blades!
Re:Hang on.. by captainclever · 2003-09-27 06:38 · Score: 2, Informative

I had to buy a new box as a pg database server recently..
almost bought a dual opteron, but chickened out and went for a Xeon instead.

the suse distro that supports it is still a bit shaky and i wanted to wait for some good bench results.

maybe my next server will be 64bit.. :)

--
Last.fm - join the social music revolution
Re:Hang on.. by FatherOfONe · 2003-09-27 06:46 · Score: 5, Informative

We have a dual Opteron 1.8GH and a 2GH on order. We found that with RedHat and Oracle the Opteron in 32 bit mode beat the crap out of a quad Xeon for the stuff we do.

Just an F.Y.I. Again this was with 32 bit code. I tried the RedHat BETA and it wouldn't even boot up without locking.

So given that Oracle cost us over $20k a processor, we saved over 40 grand!

--
The more I learn about science, the more my faith in God increases.
Re:Hang on.. by _|()|\| · 2003-09-27 07:35 · Score: 1

So given that Oracle cost us over $20k a processor, we saved over 40 grand!
I didn't know anyone actually bought processor licenses. I'm surprised that a dual Opteron beats a quad Xeon. All things being equal (i.e., 3 GHz Xeon compared to 2 GHz Opteron) I would have thought that hyper threading would give the advantage to the Xeon.
Re:Hang on.. by SD-VI · 2003-09-27 07:45 · Score: 2, Interesting

Actually, I doubt the motherboards can be tweaked all that much. You see, usually any performance increase from motherboard tweaking comes from memory controller optimization, and the ClawHammer (Athlon 64) and SledgeHammer (Athlon 64 FX / Opteron) have an on-die memory controller. I'd look more to code optimization than motherboard revisions if I were you. This also means that the performance difference between chipsets is tiny. nForce3 will probably be very popular just because of nForce2, even though it was nForce2's robust memory controller that gave it its superb reputation). Even the ALi (shudder) chipsets will do fine.
Re:Hang on.. by TheToon · 2003-09-27 07:47 · Score: 1

LOL! So you bought the hype :)

HypeThreading doesn't really buy you that much. -10% to +10% with some extremes at both ends.

--
//TheToon
Re:Hang on.. by BlueBiker · 2003-09-27 07:57 · Score: 1

Intel's forthcoming Prescott will have improved HT, will be interesting to see how much it helps on current and new applications.
Re:Hang on.. by BlueBiker · 2003-09-27 08:05 · Score: 1

That's all true, but different chipsets have different speed HyperTransport connections, which may have some impact depending on what the system is doing. IIRC, the AGP interfaces may need some work.
Re:Hang on.. by Anonymous Coward · 2003-09-27 08:27 · Score: 1, Informative

I'm surprised that a dual Opteron beats a quad Xeon
Hello!!!!! It is Oracle, a relational DATABASE, that the poster was talking about. I/O throughput counts unless you move some heavy computational analysis into the database. Very likely this isn't a "cpu" difference as much as "i/o" difference. Since the "i/o" (built-in memory controller) is bound to the physical CPU in the Opterons case it is difficult to pull those to apart.
Crunching through more threads isn't going to necessarily make an Oracle instance go faster if the blocks you need are "far away" from the CPU. Yes SMT helps to hide latency... but faster I/O helps hide latency too. :-) In the Opteron case in some circumstances it can do parallel I/O... each CPU with its own controller can pull in parallel. As long as the problems the are working on are segmented it is likely that they get twice the bandwith any one of the those 4 Xeons will get. Under those circumstances the delta in GHz, unless quite large, isn't going make much of a difference.
Opterons are nice because they allow NUMA boxes to be constructed with almost commodity parts.
Re:Hang on.. by WoTG · 2003-09-27 08:30 · Score: 4, Informative

I don't have any firsthand experience with this stuff, but I have read a lot of the reviews and whatnot. I gather that the difference is in scaling to 2 way (and 4 way) boxes. While a single 3GHz Xeon and a single Opteron are pretty competitive, the Opterons are built for dual and quad processor work - especially with memory intensive applications.

Each Opteron has it's own memory channels in multiprocessor boxes. All memory is still shared throughout the system, it's just that there is more total memory bandwidth to go around as you add Opterons. In comparison, Xeon systems have the same amount of system wide memory bandwidth from 1 CPU all the way to 4 CPU's. The net result is that in many cases a second Opteron processor nets a gain 80% or more performance - which is a LOT better scaling than Xeons. This will probably be even more evident in future comparative reviews of quad CPU boxes since the Xeons will be sputtering on memory bandwidth fumes (relative to the Opterons).
Re:Hang on.. by Gherald · 2003-09-27 08:36 · Score: 1

No it will not be interesting. It will be increadibly boring given all the marketing hype. "Improved Hyperthreading" !

Just what we need.

--
The unofficial /. digest
Re:Hang on.. by Kris_J · 2003-09-27 11:42 · Score: 1

-1 Redundant.
What you say is assumed for the first set of reviews. How long do you think we should wait? The answer is that we shouldn't, but we should be open to revising our opinion as new information and better testing arrives.
Re:Hang on.. by fitten · 2003-09-27 12:44 · Score: 1

NUMA has its bad cases as well. It's pretty easy to bring an O2K to its knees if you do stupid things (or do things on purpose to bring it to its knees).
Re:Hang on.. by Jeff+DeMaagd · 2003-09-27 14:11 · Score: 2, Interesting

You know, I can't disagree. Moving the memory controller must have greatly simplified the chipset design. Even the multiprocessing is greatly simplified as well as the CPUs are the ones that handle that, one may be able to use the exact same chipset for both dual and single chip board designs, possibly even quad designs. The down side is that a lot of dual CPU and definitely all new quad CPU systems need greater I/O, that which the single CPU designs rarely need, such as 64 bit PCI slots, at 66, 100 and 133 MHz speeds.
Re:Hang on.. by madprof · 2003-09-28 03:54 · Score: 1

That's a fair point.
Re:Hang on.. by FatherOfONe · 2003-09-28 10:02 · Score: 1

Most of our databases fit in to RAM. I do agree with what you are saying in general, however in our case it isn't true. The two biggest factors that we have is the amount of on die cache and the FSB speed. Now, as mentioned by other posters, the Opteron handles dual processor to RAM far more efficently than a dual Xeon. Also the Xeons that it is being compared against only has 512k of on die cache.

The core question that has to be asked is if you need dual processors at all. If not then a Xeon or new P4 would probably be your best bet, but if you do then the dual Opteron is a great buy for processors. Take this plus the fact that you can easily get one of these boxes with 4 GIG of RAM now, and disk i/o becomes less of a factor. Heck, if you want to throw some more money at it, you can actually get one of these boxes with 12 GIG of RAM (If someone makes 333FSB 2GB DDR chips). With 12 GB or RAM that would hide some serious disk i/o issues.

--
The more I learn about science, the more my faith in God increases.

anandtech by adpowers · 2003-09-27 06:09 · Score: 1

Don't forget Anandtech.

Pentium 4 Emergency Edition by Master+Bait · 2003-09-27 06:12 · Score: 1, Funny

It was either the Inquirer or The Register that had an interesting article saying that these CPUs (which are MP Xeons) still have their multi-cpu support enabled, thus saving astute customers thousands of dollars over their full-priced, $3,900 Xeon counterparts.

--
"Only in their dreams can men truly be free 'twas always thus, and always thus will be."
--Tom Schulman

Re:Pentium 4 Emergency Edition by MarcoAtWork · 2003-09-27 06:24 · Score: 3, Interesting

while it's very likely that this is the case for the reviewing samples (probably rushed out as fast as possible, I can't believe that Intel had plans for the EE all along given that there were 0 leaks) given that this is a paper launch by the time the EE will actually be available to customers I'm sure the MP support will disappear.

--
-- the cake is a lie
Re:Pentium 4 Emergency Edition by Unoriginal+Nick · 2003-09-27 06:25 · Score: 3, Informative

It was either the Inquirer or The Register that had an interesting article saying that these CPUs (which are MP Xeons) still have their multi-cpu support enabled, thus saving astute customers thousands of dollars over their full-priced, $3,900 Xeon counterparts.
The multi-cpu support may or may not be still enabled, but the P4 EE has a different pin count than the XeonMP, so you wouldn't be able to use it in the XeonMP boards anyway.
Re:Pentium 4 Emergency Edition by Gldm · 2003-09-27 06:34 · Score: 1

Yes, now all you need is some multi-socket 478 motherboards.... seen any?

--
Introducing the new Occam Fusion! Now with sqrt(-1) fewer blades!
Re:Pentium 4 Emergency Edition by BlueBiker · 2003-09-27 08:22 · Score: 2, Interesting

Makes ya wonder whether a company such as Powerleap might come out with a CPU adapter to support it. For a long time the Athlon MP series offered the only affordable SMP solution, especially if like me you found a pair of Athlon XPs which worked happily in SMP mode.
Re:Pentium 4 Emergency Edition by Troll_Kamikaze · 2003-09-27 16:21 · Score: 1

probably rushed out as fast as possible, I can't believe that Intel had plans for the EE all along given that there were 0 leaks

A lot of the reviewers' P4 EEs appeared to be labelled with a magic marker, so I think your suspicion is correct.

by the time the EE will actually be available to customers I'm sure the MP support will disappear

Yes. Otherwise, potential customers wouldn't buy any Xeon MPs. A 2.0 GHz Xeon MP with 512KB L2 cache and 3MB L3 cache (like the P4 EE) costs about $3,200.00. And that's at 2.0 GHz.
Re:Pentium 4 Emergency Edition by Bored+Huge+Krill · 2003-09-27 16:34 · Score: 1

not so. The P4 EE doesn't have ECC. Not that good for a server...
And I very much doubt that the multi CPU support will go away that fast. Spinning a chip and getting it validated takes a very long time indeed. ECC can be disabled with a bond-out change (not that big a deal - doesn't need a silicon spin) but multi CPU support probably can't be disabled in the same way
Re:Pentium 4 Emergency Edition by inquisitor · 2003-09-28 03:41 · Score: 1

Interestingly, Intel also supplied the P4EE clock-unlocked, so that reviewers could OC it as a spoiler for the Athlon 64.

It will almost certainly disable both the clock-lock and the SMP capability before shipping; a very Intel thing to do. A64 scares them hard. Good.

Re:great by ColdGrits · 2003-09-27 06:12 · Score: 1

so how long until Sun release an OS for 64 bit?

Oh, wait., they already did, decades ago.

What was your point agian?

--
People should not be afraid of their governments - Governments should be afraid of their people.

do they use any 64 bit applications ?? by zymano · 2003-09-27 06:13 · Score: 2, Informative

The OS is 64 but most of the games/applications are 32 bits ?????

Re:do they use any 64 bit applications ?? by lederhosen · 2003-09-27 06:29 · Score: 1

Yes!

Windows (beta) is 64 bit and can run 64 bit
applications as well as 32 bit ones.

That is, you got all your old apps and can get
new 64 bit ones that you need.

Of course there is 64 bit GNU/Linux and BSD for
it as well.
Re:do they use any 64 bit applications ?? by telstar · 2003-09-27 19:07 · Score: 1
"The OS is 64 but most of the games/applications are 32 bits ?????"
- If you build it ... they will come...

don't bother with the FX yet by Tumbleweed · 2003-09-27 06:13 · Score: 4, Informative

The socket format will be changing soon, and once the upcoming changes happen, things will be much better. You'll then be able to use non-ECC memory, and the motherboards will be less expensive.

Until then, yeah, the FX is freaking fast, but waaaaay overpriced, so don't bother.

Re:don't bother with the FX yet by Tumbleweed · 2003-09-27 06:27 · Score: 2, Informative

I think that the new Socket 939 format will be available 6 months or so before PCI Express x16 becomes mainstream. I've been reading that the switch to PCI Express won't be happening, for the most part, until the 2nd half of next year, whereas Socket 939 should be at the beginning of 2Q next year. Of course, that assumes that everything happens on schedule, which never happens. :)

It's all a moot point for me - my next computer will be a Mac G5. Because of the coming PCI Express train, I'm not going for the big duallie for now - I'm gonna get the single proc 1.8GHz machine, then upgrade to the then-biggie once they migrate to PCI Express.
Re:don't bother with the FX yet by BlueBiker · 2003-09-27 07:15 · Score: 1

The socket will be changing to 939pin for next generation dual memory channel 64-bit Athlons but not for Opterons, which the FX-51 essentially is [~Opteron 148 IIRC]. If what you want is a system upgradeable to future Opterons, starting with an FX-51 may be a relatively inexpensive approach.
Re:don't bother with the FX yet by Tumbleweed · 2003-09-27 07:19 · Score: 1

Hey, there will always be exceptions to any rule. A Hummer will be the best vehicle for some people, just not for the vast majority of people. I refuse to put such silly-ass disclaimers on all my Slashdot posts, despite the inevitable replies calling for them.

If you've got the money to burn on a system that has already been announced as an 'end of life' platform, by all means, go ahead.

A good alternative - send the money to me. :)
Re:don't bother with the FX yet by Tumbleweed · 2003-09-27 07:22 · Score: 1

The FX is a crippled Opteron - it has fewer HT links. If you want an Opteron, get an Opteron.
Re:don't bother with the FX yet by BlueBiker · 2003-09-27 07:37 · Score: 1

An Opteron 2xx or 8xx has more functioning coherent HT links, but the FX and Opteron 1xx each have one. I don't think there's any such thing as an FX motherboard, they just run in single Opteron motherboards with or w/o AGP.

People should buy what they can afford that fits their needs. If an FX-51 with ECC memory satisfies their requirements, then there's no point paying more for a 'real' Opteron.
Re:don't bother with the FX yet by _|()|\| · 2003-09-27 08:03 · Score: 1

If you want an Opteron, get an Opteron.
Except that there is no 2.2 GHz Opteron. For most users, the Opteron and Athlon 64 FX are both over priced. For those with money to burn, the Athlon 64 FX is arguably the fastest desktop processor available. Of course, if you've really got money to burn, why not get two 2.0 GHz Opterons?
Re:don't bother with the FX yet by Tumbleweed · 2003-09-27 08:52 · Score: 1

> course, if you've really got money to burn, why not get two 2.0 GHz Opterons?

Or four. Or eight.

Or an 1100-node G5 cluster! :)

I'm still waiting on some Athlon64 vs G5 comparisons. C'mon, benchers, get with it! :)

I just read that the Intel P4EE CPUs are 'accidentally' SMP enabled. Hehe. Whoopsies. Still way overpriced and I don't even wanna think about the cooling requirements, though. :(
Re:don't bother with the FX yet by Tumbleweed · 2003-09-27 09:16 · Score: 1

Wow, the "double-wide" mobo concept. That's...unique.

It doesn't really bother me, as I'm jumping ship to Apple when I can afford it. *shrug*
Re:don't bother with the FX yet by Tumbleweed · 2003-09-27 09:40 · Score: 1

> When they switch sockets to 939 pin can I still use ECC memory if I want?

Hmm, good question. I'm _guessing_ the answer will be yes, depending on mobo support. But, if you want to use ECC memory, you might as well just stick with an Opteron, and the 940 pin socket.

Personally, I'd rather have a 939-pin Athlon64 FX with dual-channel DDR400 unbuffered Mushkin 2-2-2 memory. The best quality memory you can get, and all the speed you can hope for. Oh yeah.

This is the sad thing about the Mac G5 - the mobo is apparently not designed to make use of higher-quality RAM (ie: lower-latency) than what they ship with. Pfeh. Fools. It's one thing to ship with 'regular' memory, but to hobble the machine so crazy people like me can't upgrade is _really_ irritating. Oh well.
Re:don't bother with the FX yet by Tumbleweed · 2003-09-27 10:17 · Score: 1

> The last DIMMs I bought--brand new--were so bad Linux wouldn't even boot.

Yeah, that's what happens when you buy cheap-ass RAM. Buy the good stuff (Corsair or Mushkin, if you can afford it), and rest easy.

> And speaking of future, does Athlon64 FX support DDR2?

Nope. The memory controller is now on the die with the CPU, so once AMD decides to go DDR2, they'll have to update the CPUs themselves with a DDR2 controller. Oh goodie.
Re:don't bother with the FX yet by Tumbleweed · 2003-09-27 10:41 · Score: 1

I didn't say it did, but buying new RAM that is immediately crappy is generally a sign of buying _crappy_ new RAM, not high-quality RAM.

re: "Gamer RAM"

Negligible? That depends entirely on what you do with your computer. Low-latency RAM (very different from what I think you're referring to) can have a dramatic effect in certain computing tasks. Gamers seem to go for PC4200 & up type stuff for use in overclocked situations, which is not what I'm referring to. They never pay attention to the latency of the RAM they use, because for them, it really _doesn't_ generally matter. I'd rather have 2-2-2 PC3200 over 3-6-6 PC4200, easily. But that's just me. :)

As far as I know, Mushkin is the only one offering 2-2-2 PC3200 modules at the time. Even Corsair can't match that. But boy, do you ever pay for it! (> $200 per 512Meg module *ouch*)
Re:don't bother with the FX yet by sigwinch · 2003-09-27 11:06 · Score: 1

Yeah, that's what happens when you buy cheap-ass RAM. Buy the good stuff (Corsair or Mushkin, if you can afford it), and rest easy.
Even the best RAM has, on average, 25 errors per gigabyte per year. (Got that number from Corsair's website BTW.) Awhile back one of the benchmarking sites (Anandtech?) tweaked Memtest86 to measure single event upsets and found appalling failure rates, and that was with reputable RAM, mobos, and power supplies.
If you are not using ECC memory, you have made a conscious decision to reject both correctness and security.

--
--
Kuro5hin.org: where the good times never end. ;-)
Re:don't bother with the FX yet by Jeff+DeMaagd · 2003-09-27 13:19 · Score: 1

If you are not using ECC memory, you have made a conscious decision to reject both correctness and security.

True, the same can be said of overclocking as well, but they seem to dismiss it without having real long term experience with it. When frame per second is the goal, then it is easy to ditch ECC.

If the only use of the computer is web surfing and gaming, then there is little justification to pay the extra.

I've had enough experience with (and without) ECC to say that you are right. The most stable computers I've owned had ECC. The systems with lesser stability did not. I think it is worth the extra, but then I like having solid computer first, because I do use the computer to do paying work that help pay for the thing, and I can't loose it.
Re:don't bother with the FX yet by TheLink · 2003-09-27 18:24 · Score: 1

Actually once you start having lots of RAM, ECC may become almost compulsory.

If you have 128MB the chances of one bit becoming incorrect is rare. If you have 4GB it starts becoming more likely, also manufacturers may be starting to get close to the safe limits of their process with the high capacity chips.

If ECC becomes more common it'll become more affordable. It won't be as cheap in the near future.

The big problem for AMD I see is, even if you want to buy the stuff you can't.

Once I stop seeing "prebook" in this pricelist (or similar ones) then I'd say AMD has made it past the bend at least where I am.

http://www.lowyat.net/pricelist/html/compuzone.h tm

USD1=RM3.8

These are one of the very many independent small shops that sell the beige/white boxes that make up like 30-40% of the worldwide PC sales.

Even if Dell and HP don't sell Athlon 64 based PCs, if the independents start building lots of them, AMD can survive.

http://www.itworld.com/Comp/1183/030707whitebox/

"Market research company IDC pegged the percentage of white boxes sold in the U.S. at 36.4 percent of the total market, or 4.3 million units in the first quarter. Gartner Inc. thought the amount of white boxes sold in the U.S. was even higher in the first quarter, at 4.5 million units or 38 percent of the total market. Worldwide, more than half of all PCs sold are white boxes, according to both companies."
--
- Too many replies beneath your current threshold
Re:don't bother with the FX yet by BlueBiker · 2003-09-28 15:47 · Score: 1

Well, I spoke too quickly. There're conflicting assertions all over the web, but according to AMD's website:

All Opterons have three HyperTransport links. One of the 2xx series' HT links is coherent, allowing it to talk to another CPU. All three of the 8xx's are coherent, while none of the 1xx's are. The Athlon64 and AthlonFX each have one HT link, which is non-coherent.

Quote from the article by nick-less · 2003-09-27 06:14 · Score: 1

A 64-bit computing experience means that our PCs will no longer be limited to addressing only 4GB of RAM. In a 64-bit environment, it's now possible to address 8 terabytes of memory

sound's like SCO math to me...

Re:Quote from the article by TCM · 2003-09-27 06:22 · Score: 1

Should be more like

$ echo "2^64/1024/1024/1024/1024" | bc 16777216

16M TB, they're only off by a factor of 2097152.

--
Of course it runs NetBSD. BTC: 1NT7QvbetmANwaMzhpVL6
Re:Quote from the article by The+One+KEA · 2003-09-27 06:24 · Score: 4, Informative

Not quite.

The AMD64 core uses a 40-bit physical memory address space, which is 1 Terabyte. It also uses a 48-bit virtual memory address space, which is 256 Terabytes.

A full 64-bit physical memory address would allow for 16 Exabytes of memory.

--
SCREW THE ADS! http://adblock.mozdev.org/ Proud user of teh Fox of Fire - Registered Linux User #289618
Re:Quote from the article by addaon · 2003-09-27 14:24 · Score: 1

But the x86-64 architecture is designed to allow both to scale up, as necessary, to 64 bits, without requiring program changes, yes?

--

I've had this sig for three days.

you can get more ... by flex941 · 2003-09-27 06:15 · Score: 1, Informative

Follow this link ... 2.8Ghz Athlon FX for talk and benchmarks.

P4 Emergency Edition looks like from past centruy in light of this. Ok, probably one can overclock that chip too.

Re:you can get more ... by The+One+KEA · 2003-09-27 06:20 · Score: 2, Informative

Damn! You stole my thunder; I was going to post this benchmark.

I think the biggest con of the FX51 is that soon it will be orphaned because of the 940pin -> 939pin change, which will allow that particular core to use normal DDR400 memory instead of registered ECC DDR400 memory.

--
SCREW THE ADS! http://adblock.mozdev.org/ Proud user of teh Fox of Fire - Registered Linux User #289618
Re:you can get more ... by flex941 · 2003-09-27 06:24 · Score: 1

Don't know. It was explained that 939 pin socket cpus need 4-layer mobos. And 940 pin cpus 6-8 layer mobos. Manufacturing of 4-layer ones should be cheaper.

But given the price of this particular cpu I must conclude that those extra $$ doesn't matter.
Re:you can get more ... by Sangui5 · 2003-09-27 08:05 · Score: 1

P4 Emergency Edition looks like from past centruy in light of this. Ok, probably one can overclock that chip too.

Except that the P4's overall performance is more dependant on the memory subsystem than other chips. The L3 cache of the P4EE may help it scale well at higher clock speeds, but I doubt it. It's only 2MB, and it is still being fed off of the same speed memory bus with equally poor latancies no matter how much you overclock it.

It is interesting to see how much the improved memory subsystem of the FX version lets it beat the 3200+ version hands down. High clockspeed helps, but I think that the memory improvements represent a performance boost which will scale for quite some time.
Re:you can get more ... by RzUpAnmsCwrds · 2003-09-27 15:36 · Score: 1

Solution: AMD Opteron/Athlon64 in 64-bit mode
Alternate Solution (Even Faster): Intel Itanium 2 with ICC/IFC compiled code

The P4 is fast if you have code that doesn't branch. That basically means 3D rendering and media encoding. Everything else is faster on Athlon 64.

Betas Of Athlon64 Optimized Linux by Anonymous Coward · 2003-09-27 06:18 · Score: 5, Informative

Are avalible from SuSE, Gentoo, and Debian!

SO if you are complaining "theres no 64 bit os yet", stop complaining, leave the evil empire behind and see the REAL power of opensource.

Re:Betas Of Athlon64 Optimized Linux by Brian+Stretch · 2003-09-27 11:36 · Score: 2, Informative

And Mandrake and Red Hat Enterprise (somewhere).

64-bit Windows beta is available via MSDN if you need Evil Empire compatability.

There's even a bootable CD of 64-bit America's Army. Linux based, of course.
Re:Betas Of Athlon64 Optimized Linux by tshak · 2003-09-27 12:00 · Score: 1

stop complaining, leave the evil empire behind and see the REAL power of opensource

What, that there's BETA's available? WinXP 64bit is available as a BETA as well. There's still no stable 64bit OS available yet. And none of this has anything to do with OSS.

--

There is no longer anything that can be done with computers that is nontrivial and clearly legal. -- Paul Phillips

No problem by jbarket · 2003-09-27 06:20 · Score: 1

Sure! Just let me whip out my abacus and do a couple of calculations and...

--

-----
jonathan barket

Re:Finally! by yoshac · 2003-09-27 06:20 · Score: 2, Informative

None of the above. It was Digital, with the Alpha AXP chips. It runs native 64bit Unix/VMS and 32bit Windows. Also x86 emulation via the !FX emulator.

Yeah. by Gldm · 2003-09-27 06:21 · Score: 1

Do the math! (Jaguar)

--

Introducing the new Occam Fusion! Now with sqrt(-1) fewer blades!

Re:vs. Pentium 4 AGAIN???? by jrockway · 2003-09-27 06:23 · Score: 1

Actually, I'd like some Linux-centric benchmarks. I don't care how long WinRAR takes; it's useless. What about bzip2? And then there's the kernel compile, games, etc. Before I buy something, I want to know how it performs on things that _I_ do.

That said, I have a 3000+ right now (er... a 2500+ "unlocked" to a 3000+ :-D) and won't be upgrading soon. It's fast enough and I can't use more than 512MiB of RAM at a time anyway (all programs in memory + disk cache of them are only about 300MiB for me; gcc uses some more...). I'm saving my money for a 2.0TiB RAID array :)

--
My other car is first.

Try reading the links? by Gldm · 2003-09-27 06:25 · Score: 1

The Ace's review does some testing on the already pretty useable windows64 beta. I don't think it'll take that much longer for it to get up to their normal stability standards for a release.

--

Introducing the new Occam Fusion! Now with sqrt(-1) fewer blades!

Re:Try reading the links? by Anonymous Coward · 2003-09-27 07:23 · Score: 1, Funny

If it's a beta it's already up to their normal stability standards.

(Someone had to say it)
Re:Try reading the links? by eatdave13 · 2003-09-27 15:29 · Score: 1

You're a idiot.

--
"Verbing weirds language." -- Calvin

these chips are beasts by ghettoreb · 2003-09-27 06:25 · Score: 1

this is a godsend for the people that like to brag about the number of opterons they have in their gaming PC, but these chips are too darn expensive for the majority of us. I wish that AMD will create some low end chips in the $150-$200 range. The AthlonXP's they have out now in that range are too slow for their price.

I would also like to see motherboards become cheaper because chipsets no longer have to have a memory controller (since the A64's have one built in)

just my $.02

Re:these chips are beasts by levik · 2003-09-27 07:06 · Score: 1

I agree with you... I would like to see motherboards and ram become cheaper. Also I want processort to become cheaper, and monitors could stand being a little cheaper than they are.
Other things that should be cheaper are electricity and internet access. Laptops should definitely be A LOT cheaper, as should air travel.

--
Ñ'
Re:these chips are beasts by minion · 2003-09-27 07:09 · Score: 1

I wish that AMD will create some low end chips in the $150-$200 range. The AthlonXP's they have out now in that range are too slow for their price.

I would also like to see motherboards become cheaper because chipsets no longer have to have a memory controller (since the A64's have one built in)

So essentially, you're a real cheap-ass and want something for nothing. Maybe you could send your congressman a letter suggesting welfare for geeks.

--

-- If we don't stand up for our rights, now, there will be no right to stand up for them later.
Re:these chips are beasts by spinlocked · 2003-09-27 07:18 · Score: 1

...I wish that AMD will create some low end chips in the $150-$200 range...

Why? Do you have >4GB of RAM?

--
# init 5 Connection closed.

Oh... ...bugger.

Re:vs. Pentium 4 AGAIN???? by notanatheist · 2003-09-27 06:27 · Score: 1

at least Tomshardware.com threw in the results from earlier Opteron tests for comparison in his full blown review. Seems the Athlon still holds out in processes that require quicker execution while the P4 is still best for anything CPU intensive. Another thing, how reliable are the new AMD chips going to be? AMD should not allow certain motherboard makers to have boards for their chips. There's only a few reliable manufacturers out there yet people still buy the cheapest board, ram, and power supply and complain about how your system is unstable. /end rant

I, for one, by civilengineer · 2003-09-27 06:28 · Score: 1

welcome a plethora of processor reviewers

--

New year Resolution: Don't change sig this year

Re:vs. Pentium 4 AGAIN???? by Gldm · 2003-09-27 06:30 · Score: 1

Yeah, it's not like you can just look them up. Oh wait, Apple doesn't like to send them scores. Wonder why that is....

--

Introducing the new Occam Fusion! Now with sqrt(-1) fewer blades!

Re:64bit.. Schmobit... by notanatheist · 2003-09-27 06:31 · Score: 2, Funny

this is not a problem with an abacus as you can own an original 8 bit abacus and just add 8, 16, 32, or even 64bit upgrade kits. You may need a larger frame to accomadate the beads but you'll have years of growth from your very own abacus.

Re:Finally! by Gldm · 2003-09-27 06:36 · Score: 1

I seem to remember a few DEC Alpha based "PC"s running WindowsNT back in the early 90s. Does that count?

--

Introducing the new Occam Fusion! Now with sqrt(-1) fewer blades!

Got me thinking by BiggerIsBetter · 2003-09-27 06:38 · Score: 1

I wonder when/if Sun will release a 64 bit Solaris x86?

--
Forget thrust, drag, lift and weight. Airplanes fly because of money.

more motherboard reviews, please (esp. w/ Linux) by _|()|\| · 2003-09-27 06:38 · Score: 3, Informative

I haven't made all the rounds, but it seems like everyone is using the same two motherboards: Asus SK8N and MSI 8KT. I really like the looks of Monarch's Hornet 64, with a uATX Gigabyte GA-K8VT800M. I'd like to see some reviews first, especially regarding chipset support under Linux. I'd also like to hear more about video drivers. I've heard that NVIDIA's drivers need some work. (Does ATI even have any?)

We've got a couple of Opterons at work, one for 32-bit compatibility testing, and another for the AMD64 port. It's pretty cool to see this in Python on SuSE Linux 8.2 beta:

>>> type( 9223372036854775807 ) <type 'int'>

SuSE Linux 9.0 for AMD64 is supposed to ship next month. Hopefully, it will be a little cheaper than RHEL 3.0 for AMD64, which will be more than twice the price of RHEL 2.1 for x86!

Establishing the ultimate randomness by Qbertino · 2003-09-27 06:39 · Score: 1

Slot, Socket, Boardfrequency, Memory Type, CPU Bitcount and CPU Class...
Bit by bit (no pun intended) vendors are establishing a true real life randomness of standards. A shure sign that computers are becoming a comodity. Soon we'll see the same with operating systems.

--
We suffer more in our imagination than in reality. - Seneca

Proper benchmarks by NitroPye · 2003-09-27 06:39 · Score: 3, Insightful

Untill I see proper benchmarks done with applications built for 64bit chips I am going to go pffffft.

I am still blown away that the FX did better then then 3.2 P4.

Proper benchmarks include not using a 64bit beta stupid o/s like windows, a properly optimized linux (suse 64 or gentoo) and applications built for the chip. Openoffice, kde and kde apps, mozilla, some miscellaneous 3d engines running some impresive demos, maybe tenebrae quake. Tenebrae quake is great being that its open source and takes a huge amount of gfx and proc power.

"organic" ICs by Animats · 2003-09-27 06:49 · Score: 1

What's this "organic" stuff? Apparently, plastic IC packaging (as opposed to ceramic) is now referred to as "organic".

Re:"organic" ICs by cosmo7 · 2003-09-27 07:01 · Score: 3, Funny

Organic integrated circuits are produced without using fertilizer or pesticide.
Re:"organic" ICs by oliverthered · 2003-09-27 07:12 · Score: 1

Organic IC are made from sylicone waffers that have been grown from.....

I expect that they use some kind of thermoset plastic (organic in chemistry) and not a ceramic.

--
thank God the internet isn't a human right.
Re:"organic" ICs by Minna+Kirai · 2003-09-27 14:50 · Score: 1

The word organic means "composed of carbon, hydrogen, oxygen, and nitrogen". Petroleum & plastic fits that definition.

There may be some confusion with a recently-invented marketing term, "Organic", which applies to agricultural products grown with sophistication below some arbitrarily-defined level.

In reality, any food a human could survive eating is organic, by the scientific definition.

Most Only 32 Bits... by MBCook · 2003-09-27 06:52 · Score: 4, Informative

The reviews are basically all 32 bit, which is a shame. Linux is out there. I've only seen one review that did anything really 64 bit (running win32 programs on win64 doesn't count). The only 64 bit test they did under Linux was MP3 encoding. The test was the Athlon 64 running a 32 bit version they compiled of the MP3 encoder vs running a 64 bit version of the same program. The "bitness" was the ONLY thing that was changed.

The results? The 64 bit version took nearly HALF THE TIME of the 32 bit version. This is the kind of thing we have to look forward to in some things (MP3s, video encoding, etc).

The Athlon 64 is fast in 32 bit mode, and can beat a P4 many times. But when the 64 bit code comes along, the P4 will be taking one hell of a beating.

PS: Sorry I don't remember which review had this test. I don't have time to go hunting for it right now.

--
Comment forecast: Bits of genius surrounded by a sea of mediocrity.

Re:Most Only 32 Bits... by TrancePhreak · 2003-09-27 07:27 · Score: 1

You forget that the P4 with an optimized MP3 encoder (IE not LAME) beat out the AMD chips even at 1.8gHz before the 64FX came along. The extra registers and such help, but not as much as some good ol' SSE2 optimizations with a compiler that does all the work for you.

--

-]Phreak Out[-
Re:Most Only 32 Bits... by BlueBiker · 2003-09-27 07:28 · Score: 1

It's my understanding that while AMD64 chips give a nice boost in integer calculations [largely due to extra GPRs], their SSE2 floating point is only about on par with the fastest P4s. Later versions should have improved SSE2.
Re:Most Only 32 Bits... by Sangui5 · 2003-09-27 07:56 · Score: 4, Insightful

The test was the Athlon 64 running a 32 bit version they compiled of the MP3 encoder vs running a 64 bit version of the same program. The "bitness" was the ONLY thing that was changed.
Not really. While the "bitness" changes, what also changes is the number of registers visible to the compiler. The x86 ISA has been dealing with internal register rename as a nasty hack to deal with a sever shortage of programmer visible registers for a long time. This goes to show that the compiler is much smarter about register allocation than a hardware renamer can ever be. I'm interested in seeing the performance of common multimedia applications once hand-written core loops are available.
And a note to those who are pointing to improved SSE2 support as the reason for the performance gain: they are comparing an AMD64 in 32 bit mode vs one in 64 bit mode. Unless GCC is being bass-ackwards, the SSE2 support should be benefiting the 32 bit mode as well. It appears that the only variables in this benchmark are the 64 bit math and the additional registers.
Re:Most Only 32 Bits... by Jeff+DeMaagd · 2003-09-27 08:51 · Score: 1

a ~30% increase in speed isn't "half the time", but it is a nice performance increase, and being only one test, it isn't enough to work from, and a more valid test would compare it to P4 running Linux too. Given that media encoding under ia32 is one of the weaker points of the Hammer architecture, then that means it can catch up and possibly beat the P4 here.

In a way it is a shame that the alleged hardware review sites really don't check out Linux in part of their testing. Linux stands to score a major publicity coup, although in reality I doubt gamers really want to deal with Linux yet unless there is something on it they want on it, as XP64 for AMD64 is coming out early next year.

Re:great by kyrre · 2003-09-27 06:53 · Score: 1

Linux has been availible for 64-bit plattforms for years as well.

64 bit resources by JusTyler · 2003-09-27 06:54 · Score: 2, Insightful

I've been struggling to find any good 64 bit resources for Linux.

Basically, I want to know about all the 64 bit versions of major apps and systems, like MySQL, Perl, and so on. I know Perl is in 64 bit, because you can compile it to be, but what about stuff like MySQL, Apache, TomCat...

Post your best 'going up to 64 bits on Linux/FreeBSD/elcheapo UNIX' resources here, and attract some karma :-)

--
Web Hosting Reviews

Re:64 bit resources by exhilaration · 2003-09-27 07:09 · Score: 1

You just recompile any app to "be 64-bit". If you want a pure 64-bit system you can do it, as long as you can also recompile all your drivers.
Re:64 bit resources by Minna+Kirai · 2003-09-27 19:39 · Score: 1

You just recompile any app to "be 64-bit".

That may not always be true. It's possible, especially with the C/C++ language, that a programmer will have inadvertently built in assumptions about data types.

For example, a program my try to read 4 bytes from a disk file into an "int". On a common modern CPU, that works fine. With an Athlon64 it'd leave half of the integer unfilled.

64 bit: necessary? by Dan+Farina · 2003-09-27 06:55 · Score: 1

Other than the extra memory addressing, I don't see why 64 bit is really such a big deal.

Granted, it doesn't hurt anyone, but the AMD 64-bit processors have many other fine engineering tweaks to make them as fast as they are. I suspect that the main reason for the performance increases in 64 bit mode are (and someone correct me if I'm wrong here) the extra registers that open up in this mode.

So you can address 2^64 bytes of memory, but no one loads them a word at a time anyways, it's all paged over in chunks to predict what you're going to need in registers next.

The main gains, I suspect, are in simply high IPC (which boils down to a lot of optimizations and transistors) and notably the on-die uber low latency (indeed, practically no latency) memory controller. But these don't make a big marketing splash, nor can they be simplified to a nice rally cry. Hell, most of us don't know the entirety of the specific optimizations, but do know that smiply being able to address a bigger word doesn't mean much to most, and thus that the chip can stand on other merits alone.

This is a great chip (once costs get down) but I don't see the rally cry of 64 bit to really mean anything except to servers, to the consumer it will be mostly useless and obsolete by the time we need such vast amounts of memory. (Then again, you never know how far AMD will stretch their cores, which is a very good thing, good motherboards two years ago could run most of the chips up until present day with an update.)

Am I half-correct, delusional, or just plain misinformed and my basic knowledge of machine structures utter crap?

Re:64 bit: necessary? by Dan+Farina · 2003-09-27 07:24 · Score: 1

but really, look at most of these optimizations, most of them have nothing to do with the intrinisic ability to handle 64 bit words.
Re:64 bit: necessary? by umeboshi · 2003-09-27 17:49 · Score: 1

but what about mmx,sse?
if there are also 64bit counterparts of theses sets,
the data throughput should be doubled on each register. If also, there are more available registers, that's even more throughput.

--disclaimer
I don't know much about x86,
my last assembly program was on a 6502

Re:Most Only 32 Bits... FOUND IT by MBCook · 2003-09-27 06:55 · Score: 4, Informative

Found it! It was Anandtech. Check out the bottom of the 32 bit vs 64 bit page of the review.

--
Comment forecast: Bits of genius surrounded by a sea of mediocrity.

Re:vs. Pentium 4 AGAIN???? by FatSean · 2003-09-27 06:59 · Score: 1

Shhhhh.....you'l get moderated as a troll....

--
Blar.

Re:more motherboard reviews, please (esp. w/ Linux by Drathos · 2003-09-27 07:11 · Score: 1

According to amazon.de, the 64-bit edition of SuSE 9.0 is the same price as SuSE 9.0 Pro (around $80). This means it's simply the Pro edition compiled for x86-64. I'm sure there will be a version of SLES (if there isn't already) for x86-64. That will be the one to compare to RHEL.

--
End of line..

Re:vs. Pentium 4 AGAIN???? by Lord+Kano · 2003-09-27 07:13 · Score: 1

The Athlon FX is essentially an opteron rebadged as a desktop chip just the the P4 EE is essentially a 3500 top end Xeon with a higher bus speed rebadged as a desktop chip.

Almost. The Athlon FX is an Opteron with 32 bit processing abilities. The P4EE is a Xeon with more cache.

I know some mac zealot will respond that software hasnt been optimized for the G5 yet, well it hasnt been optimized for x86-64 yet either.

And you'd both be right.

LK

--
"Hi. This is my friend, Jack Shit, and you don't know him." - Lord Kano

Re:vs. Pentium 4 AGAIN???? by l0ungeb0y · 2003-09-27 07:15 · Score: 1

Holy crap.

Gee, sorry to have pissed on your cheerios.
You know there are some of us who can give a rats ass about the P4.

There are some of us who'd like to see the difference between all 64bit processors out there.

I'm sorry that some of you can't wrap that around your pointy little heads, but I digress.

Look, IMO, benching a 64bit CPU vs a 32bit CPU while interesting is overkill when all the sites do basically the same review.

While the redundancy of these reviews can be good for QA of the overall reviewing process, the results of the P4 vs A64 are only usable for those that are concerned over the benefits of upgrading their desktop machine/lowend server.

For those of us that are scientists, animators, et al that currently plan on implementing a 64bit solution, it would be nice to see where things stack up.

So yeah, seeing the G5 bench against the A64 or the Opteron would be more beneficial and also let us see where the AMD/IBM crop sits against Intel/HP.

Solaris on dual BOXX Opteron by BiggerIsBetter · 2003-09-27 07:15 · Score: 1

Hmmm. It would seem so...
http://www.sun.com/bigadmin/hcl/data/systems/detai ls/291.html

--
Forget thrust, drag, lift and weight. Airplanes fly because of money.

Re:Solaris on dual BOXX Opteron by sl0ppy · 2003-09-27 07:22 · Score: 1

it would make sense.

one would think that a port of solaris to 64bit x86 would be fairly trivial. they already have 64bit support in their code base, and can't have diverged the sparc/x86 source bases too much, and still keep code manageability.

plus, sun already has experience with 32bit drivers in a 64bit architecture, from when they rolled out their ultra-1's.

they could possibly just do a recompile.

Free OS's? by -tji · 2003-09-27 07:29 · Score: 1

Until now, when only the Opteron was available for AMD64 support, there was very little effort for free OS support. There were efforts by RH and Suse for expensive enterprise server OS's..

Now that the low end 64 bit chip is out, what is the best Linux Distro that is freely available, or at least cheap??

Air travel? by BiggerIsBetter · 2003-09-27 07:42 · Score: 1

It cost me .02 Euro in airfares to fly to from The Netherlands to London, and then on to Venice. Total including all fees/taxes/etc was under 50 Euro. There are heaps of cut-price airlines, at least one of which will have super-cheap flights when you need one. If you pay a lot for air travel, you're flying with the wrong airlines.

--
Forget thrust, drag, lift and weight. Airplanes fly because of money.

Re:Air travel? by arivanov · 2003-09-27 09:01 · Score: 1

Or have your humour gland amputated at birth.

--
Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
Re:Air travel? by BiggerIsBetter · 2003-09-27 09:14 · Score: 1

All of the items listed have traditionally been expensive, and the poster was sarcastically pointing out that his parent poster was being immature and unreasonable in expecting things to be cheaper just because he wants them to be. However, in his rush to clever he took a service which is often perceived to be expensive as an example. He was incorrect in doing so, and I was pointing out to others who may have also thought that air travel was indeed expensive, that it is in fact already inexpensive. Oh, and yes, I get the intended joke. Do you understand now?

--
Forget thrust, drag, lift and weight. Airplanes fly because of money.

Re:vs. Pentium 4 AGAIN???? by WNight · 2003-09-27 07:42 · Score: 1

Do the XP 2500+ chips overclock well? I've got one...

Re:troll by BlueBiker · 2003-09-27 07:44 · Score: 1

I'm not sure we know how fast G5 actually is, given the lack of a 64-bit OS under which to benchmark it.

KISS, I like it. by Thinkit3 · 2003-09-27 07:45 · Score: 1

The Doom authors noted how much simpler it was than Wolf3D simply because of the move from sixteen to thirty-two bit.

--
-Libertarian secular transhumanist

Re:more motherboard reviews, please (esp. w/ Linux by _|()|\| · 2003-09-27 07:50 · Score: 1

[SLES] will be the one to compare to RHEL.

SLES 8 for AMD64 is already available, but $750 is a bit steep for an Athlon 64 PC. Considering the two-year errata policy for SuSE Linux Pro, it seems fair to compare it to RHEL WS.

I actually prefer Red Hat, but it no longer has anything in my price range. The Fedora project just doesn't seem compelling.

Re:vs. Pentium 4 AGAIN???? by BlueBiker · 2003-09-27 07:53 · Score: 1

Most definitely, see here for an overclocking starting point. They're also an excellent bang/buck, currently $84usd at Newegg.

Re:great by siddhartha03 · 2003-09-27 08:07 · Score: 1

Ummmm Windows NT for 64bit Alpha? When was that mid or early 90's? Do your research... there isn't much point to being over zealous...

--
Sock puppets stole my sig.

Re:vs. Pentium 4 AGAIN???? by BlueBiker · 2003-09-27 08:12 · Score: 1

No, Athlon64 / FX / Opteron all have the same 32-bit and 64-bit processor modes.

Re:Most Only 32 Bits... FOUND IT by Schugy · 2003-09-27 08:13 · Score: 1

AnandTech is a stupid page. I can't get any info without flash. Thanks for nothing. It's on my blacklist now because of amazing stupidity. But anyway, the Athlon 64 is great. Schugy

AMD compiler by Eric+Ass+Raymond · 2003-09-27 09:05 · Score: 1

Very interesting, but what's been holding me back from acquiring an AMD CPU is that lack of a native compiler.

In comparison to GCC, Intel's compilers produce 10-20% faster code when it comes to computationally intensive tasks. That marvellous piece of software truly pays itself off on an Intel architecture.

I don't understand why AMD is not interested in producing a compiler (even by contributing to GCC) that would help them produce similar results.

--
BOO! TERRO

Re:AMD compiler by ozzee · 2003-09-27 17:16 · Score: 1

I don't understand why AMD is not interested in producing a compiler (even by contributing to GCC) that would help them produce similar results.
Simple - they don't really need to. The reason why intel needs to work on it is because of the huge pipeline in the p4. The Athlon 64 is a speculatively executed multiple issue register renaming type that does much of the hard work on schduling instructions dynamically so it will run p4 optimized code just as well as p3 optimized code. The A64 design limits the clock speed BUT it does far more useful work per cycle than a P4. Intel played the MHz game and found themselves floundering in the Itanium and had to concede that MHz isn't everything. The A64 is simply a well designed fire breathing CPU.
Re:AMD compiler by Celandine · 2003-09-28 01:57 · Score: 1

You do know that the Intel compiler produces code that's about a similar amount faster on 32-bit Athlons, too? I tried it. Athlons and P4s both get a boost from the Intel compiler. The fractional increase is marginally more for the P4s in the applications I tested (scientific computing with lots of numerical integration) but nothing to write home about. This is probably all a bit irrelevant, since 32-bit Athlons aren't worth buying any more; I agree the lack of a native 64-bit compiler is a stumbling block.

Re:more motherboard reviews, please (esp. w/ Linux by Eric+Ass+Raymond · 2003-09-27 09:31 · Score: 1

It's pretty cool to see this in Python on SuSE Linux 8.2 beta

I don't get it.

It is an integer. What else should you see?

--
BOO! TERRO

Re:more motherboard reviews, please (esp. w/ Linux by Anime_Fan · 2003-09-27 09:42 · Score: 1

I don't get it.

It is an integer. What else should you see?

Hmmm... 9223372036854775807 was longer than 32 bits the last time I saw it, so type 'long' or something? bigint maybe?

I mean 7FFFFFFFFFFFFFFF (hexadecimal for those not quite as knowing) is like the largest signed integer for 64 bits CPU's... Handling these numbers as normal integers should be a huge boon to number crunching.

Or as the usual Cray responds to ./configure
sizeof short: 8
sizeof int: 8
sizeof long: 8

I do however think the AMD-64 thinks a short isn't 64-bit yet, however,,,

Re:vs. Pentium 4 AGAIN???? by skajake · 2003-09-27 09:46 · Score: 1

The Athlon 64 is in direct competition with the Pentium 4. It is not being marketed as a server/workstation chip ala Opteron. This is why the benchmarks involve the P4EE and A64. They are the primary competition for tommorow's desktop market.

Dont confuse the Athlon64 to be just another server chip, this big boy is aimed at the home user!

--

~ Maintainer of the Skajake Projects

Re:Fucking AMD fanbois by hawkbug · 2003-09-27 10:05 · Score: 1

You must be drunk - I'm not an AMD or Intel fanboy as you call it - I just know that AMDs floating point performance crushes any intel cpu currently on the market. That is the only way AMD has been able to stay with Intel in the performance area. Intel relies on mhz scaling in their long pipeline, while AMD tries to squeeze extra performance out of the lower mhz they use.

Re:great by ColdGrits · 2003-09-27 10:13 · Score: 1

Actually, if you are going to get pedantic, then so was Windows.

Which renders the original anonymous poster's flamebait pointless.

--
People should not be afraid of their governments - Governments should be afraid of their people.

Re:more motherboard reviews, please (esp. w/ Linux by _|()|\| · 2003-09-27 10:33 · Score: 1

What else should you see?

On a 32-bit system, you'll see something like this:

>>> type( 2147483648 ) <type 'long'>

No big deal; it's just one of the first things I tried.

STOP USING 3DMARK 2003! by BrookHarty · 2003-09-27 11:12 · Score: 1

I know this is a CPU benchmark, but look at the GFX cards they use.

Every benchmark shows the ATI 9800 Pro to be faster than the FX 5900 Ultra, in every benchmarked, (3dmark2001 included) except 3DMark2003.

Using 3DMark2003 while informative, shows a negative performance compared to all other benchmarks. The raw FPS scores prove that 3DMark2003 is not giving true proformance of games out today.

Be nice when HL2/Doom3 is out, we can compare and see if 3DMark2003 is providing true numbers for features not out yet. But then, whats the use of a benchmark program that doesnt provide benchmarks for data thats currently out?

Other than that, good stats on the new AMD FX line, and with SSE2, makes up for those games favoring Intel.

Athlon FX 2.8 Ghz today! by sundling · 2003-09-27 11:18 · Score: 2, Interesting

http://www.aceshardware.com/read.jsp?id=60000253
has a cooling technology that allowed them to overclock to a 2.8 Ghz Athlon FX. It was pretty impressive stuff, especially how well Age of Mythology did, even against the non-shipping P4 Emergency Edition.

I can't wait until theres 64 bit games for this sort of thing. Of course, the first to be released will be Unreal Tournament. Oh yahh!! I know I'll own at least one of the AMD64 computers within the next year!

I'd love to have a new AMD64 for my application server. There were benchmarks on extreme tech showing a 1.6 Ghz dual opteron beating a 2.8 Ghz quad xeon by gigantic margins (almost double). So when these Athlon 64 processors trickle down to the same clocked Opterons, it should start to shake up the server market even more. Especially given upcoming backing from Sun:
http://www.theinquirer.net/?article=11772

Re:Athlon FX 2.8 Ghz today! by sundling · 2003-09-27 11:21 · Score: 1

Oops I meant it was a dual opteron beating a dual xeon by almost double (which was just a bit above double the opteron's Mhz). All those 2s, I accidently multiplied one. Given the tremendous scaling of these processors, I'm not suprised to hear someone else mention that a dual opteron beat a quad xeon. I imagine thats where the quad part came from.

not with 4 dimm slots by chipace · 2003-09-27 11:22 · Score: 1

I completely agree with you... AMD could have released an athlon xp with an integrated memory controller and had similar performance gains (all in 32bits).

I need > 4GB of memory, and was looking forward to these chips. I have yet to see a motherboard with more than 4 dimm slots/cpu. A 2GB dimm is twice the price of two 1GB dimms (I was pricing them at 2GB = $1200).

I'm glad that AMD and Apple are pushing the technology ahead... but I don't see it being price effective until after another year.

Re:Fucking AMD fanbois by MsGeek · 2003-09-27 11:47 · Score: 1

I don't think he's talking about G3 or G4 here. G5 is the one with the heavy duty FPU circuitry. The IBM version of the Velocity Engine is not as good as Mot's, but considering all the other advantages of G5 nobody with a PowerMac G5 is going to miss much. When a native version of MacOS X 64-bit comes out, watch out.

--
Knowledge is power. Knowledge shared is power multiplied.

Re:vs. Pentium 4 AGAIN???? by timeOday · 2003-09-27 12:04 · Score: 1

Not the itanium, it's in the class of the SPARC and Power4 due to its high cost.

What I really want to see is how the Athlon64 will run 64 bit code! If AMD is smart they will make sure gcc compiles *awesome* code for AMD64. They're really losing a selling point so long as the main feature is fast execution of IA32 code - Intel is pretty hard to beat at that game. They should have worked with RedHat to make sure AMD64 ran linux excellently on release.

Ugh... by wicka_wicka · 2003-09-27 12:15 · Score: 2, Interesting

Doesn't anyone realize that the Athlon 64 isn't going to be anything special until they start testing on a 64-bit platform???

--
hi

Re:64bit.. Schmobit... by utexaspunk · 2003-09-27 13:53 · Score: 1

256 beads ought to be enough for anybody...

Overheating prevention by doc+modulo · 2003-09-27 15:35 · Score: 1

I have this video of a test Tom's Hardware Guide did.

They ran Quake 3 on several CPU's and then removed the heatsink and fan while it was running.

The 32 bit AMD Athlon burned to a smoking piece of junk, also destroying the motherboard

The Pentium 4 with overheating protection kept running Quake 3 until it slowed to 1 FPS. After the heatsink and fan were put back on, QUAKE RAN AS IF NOTHING HAD HAPPENED!

Now that's a good feature to have on a PC, especially a server on a hot day or when your fan finally seizes up.

I've read somewhere that the new 64 bit Athlons have some kind of overheating protection too, sounded like the clockspeed will be throttled automatically, can anybody verify this because I was about to buy a Pentium 4 over an Athlon 64 because of this feature.

--
- -- Truth addict for life.

Re:Overheating prevention by Ratphace · 2003-09-29 04:56 · Score: 1

I dunno about you, but my BIOS shuts down the PC if the temperature gets to a threshold that I set. I mean, who cares if Q3 keeps running, it's not like you can play the game at 1fps, I would rather the PC shutdown and not take any chances so at that point I can investigate the trouble.

Just my thoughts...

p4 ee and xeon mp by john_uy · 2003-09-27 15:46 · Score: 1

the p4 ee has ecc disabled. this is a no no for servers. so even though the xeon and p4 ee are the same by design, features have been disabled.

--
Live your life each day as if it was your last.

buy cpus next year by john_uy · 2003-09-27 15:52 · Score: 1

do not buy systems this year. first off, you will be milked into buying registered mememory which is quite expensive at the moment.

and since application support for 64 bits is catching up, expect more stable releases by next year making the cpu more attractive. and if you buy next year, you will get much faster cpus than today for relatively the same price.

maybe we can wait for the .9micron process so you will be able to get the cpus much cheaper and cooler (hopefully.)

for those regular users, windows xp will be released next year so this is mostly for gurus that mostly use linux.

guess that's it. :)

--
Live your life each day as if it was your last.

Re:64bit.. Schmobit... by Waffle+Iron · 2003-09-27 16:03 · Score: 1

Um, anyone knows how many bits can an abacus counts to?

It depends on each abacus's architecture. For example, a typical toy abacus with 10 rows has 10*ln(10)/ln(2) = 33.2192 bits.

Re:Fucking AMD fanbois by 0000+0111 · 2003-09-27 16:34 · Score: 1

I don't think he's talking about G3 or G4 here.

No he's not, he's smoking crack.

The IBM version of the Velocity Engine is not as good as Mot's

We don't know that yet.

Re:more motherboard reviews, please (esp. w/ Linux by Drathos · 2003-09-27 17:10 · Score: 1

I'm willing to bet that there will be an "Update" version of SuSE 9.0 for x86-64 (I know there will be one for x86) which is only $50. The only diff between it and Pro, is that the Update doesn't have as many manuals in the box.

I've always felt I got my money's worth when buying SuSE, especially now that they make the Update editions.

--
End of line..

Hear hear - more address space please by Namarrgon · 2003-09-27 17:56 · Score: 2, Insightful

Yes, yes it is.

Some apps require large blocks of contiguous memory - and with only 2GB of address space available, you can actually run into address space fragmentation problems long before you run out of physical memory. There simply isn't a large enough span of addresses available to map the memory into.

Other things compete for address space too. System DLLs map themselves into various places, leaving too-small gaps between them. Threads reserve 1 MB each, for the stack grow. Some PCI boards (e.g. HiDef video capture) map their buffer memory into your address space for easy access - which can be as large as 512MB!

Yes, more address space is needed even more than more physical memory.

--
Why would anyone engrave "Elbereth"?

Tyan Thunder K8W - 16 GB :-) by Namarrgon · 2003-09-27 18:04 · Score: 1

The Tyan Thunder K8W mobo is the dual Opteron board to have. Perhaps not for "consumers", but certainly as a workstation board it's unparalleled.

8 DIMM slots (4 attached to each CPU, unlike every other workstation board out there) means 8 GB using cheaper 1 GB DIMMs, or 16 GB max. It also means up to 10 or 12 GB/s of total bandwidth :-) No other dually motherboard I have seen offers both AGP and memory connected to each CPU.

Add to that an AGP Pro slot, a few PCI-X slots (100 MHz and 133 MHz), 4-way SATA RAID 0+1, Gigabit LAN (connected to PCI-X), Firewire etc, and you have the most desirable mobo available today - if you can afford it.

--
Why would anyone engrave "Elbereth"?

I keep a fire extinguisher near my desk. by pr0ntab · 2003-09-27 18:34 · Score: 1

Just in case.
These Opterons get toasty when I turn off my CPU fans because I'm watching Matrix Revolutions.

j/k. Opteron (i.e. AMD64, FX) has a built-in thermal solution. It's a little late in the game, but a welcome addition.

Anyway, that video's OLD man!
And since that video, AMD got on the stick and forced mainboard vendors to implement thermal detection/CPU protection otherwise they'd refuse to certify the motherboards.

Next you'll be telling me about this new amazing "ginger" thing that's going to revolutionize city design.

--
Fuck Beta. Fuck Dice

Re:I keep a fire extinguisher near my desk. by doc+modulo · 2003-09-29 08:28 · Score: 1

Opteron (i.e. AMD64, FX) has a built-in thermal solution. It's a little late in the game, but a welcome addition.

Thanks for the info, I was thinking about the P4 because the heat protection is implemented as throttling, not shutdown but actually I'm prefer the Athlon 64 camp because, as someone else said, Intel seems arrogant.

Now that the Athlon 64/FX seem to have throttling overheat protection I'm going to buy those from now on.

The reason I'm into it is because I'd like my (future) fanless home server to survive a heatwave and throttle back up again in the evening.

And not fanless as in "just take off the heatsink and fan" but using a case with heatpipes to the giant heatsink that makes up the side of the case or the
"natural cooling system" case.

--
- -- Truth addict for life.

Talk about Athlon 64 reviews??? by Jango64 · 2003-09-27 19:14 · Score: 1

Here what my favorite AMD site has to say:

1 - AMD Athlon 64 FX-51 at Lost Circuits
2 - AMD Athlon 64 FX-51 at OCAU
3 - AMD ATHLON64 FX-51 CPU at Hexus
4 - The Athlon 64 FX-51 Processor at Hot Hardware
5 - Athlon 64 Vs. Pentium 4 at HardOCP
6 - AMD Athlon 64 FX-51 at Legit Reviews
7 - AMD Athlon 64 & Athlon 64 FX at AnandTech
8 - Athlon 64, FX and Pentium 4 Extreme Edition at Ace's
9 - AMD Athlon 64 and Athlon 64 FX at Planet 3DNow!
10 - Athlon 64 and FX-51 vs P4 Extreme Edition at Sudhian
11 - Athlon 64 FX-51 vs. P4 Extreme Edition at X-bit labs
12 - AMD Athlon64 3200+ 32/64-bit Processor - PC stats
13 - AMD's Athlon 64 processor at Tech Report
14 - AMD Athlon 64 & Athlon 64 FX at Hardware.fr
15 - AMD Athlon 64 FX-51 Launch at SimHQ
16 - Athlon 64 FX and Athlon 64 P4 Extreme at THG
17 - Athlon 64/FX vs. P4 Extreme Edition at tecchanel
18 - Athlon 64 FX 51 & Athlon 64 3200+ at AMD Zone
19 - AMD K8 - Part 2: Athlon 64 versus All - x86-secret
20 - AMD Athlon 64 3200+ & Fx-51 at HardTecs4U
21 - AMD Athlon 64 : FX-51 performance - UK Gamers
22 - AMD Athlon 64 3200+ at Hardwareluxx
23 - AMD Athlon 64 Fx-51 & Athlon 64 3200+ at Clubic
24 - AMD Athlon 64 3200+ and Athlon 64 FX-51 - AthlonXP
25 - Athlon 64 FX-51 @ 2,2GHz at Gamers Depot
26 - AMD Athlon FX-51 Processor at Sharky Extreme's
27 - Athlon 64, kladivo na Intel p?ichazi at Svet hardware
28 - Athlon 64 & Athlon 64 FX Processors at Digit-Life
29 - AMD Athlon 64 FX-51 preview at FiringSquad
30 - First Tests of Athlon 64 PCs at PC World
31 - Mini-Review: Athlon64 FX 51 at ForumPCS
32 - Athlon 64: AMD Plays Its Trump Card at ExtremeTech
33 - Athlon64 Fx-51at TweakPC
34 - The hammer - AMD again in front! at Chip
35 - AMD64: AMDs 64bit-Architektur at K-Hardware
36 - Athlon 64 Fx-51 and Athlon 64 3200+ at Computerbase
37 - AMD ATHLON 64 FX-51 at Motherboards.org
38 - Athlon 64 FX-51 at Amdmb
39 - Athlon 64 3200+ - Forgotten CPU at Amdmb
40 - AMD Athlon 64 FX-51 at Hardinfo
41 - Athlon 64 FX51 at Hardware Analysis
42 - Athlon 64 3.200+ at Hardware Upgrade
43 - Athlon 64 & Athlon 64 FX at Hardware Upgrade

AMDBoard

Increased performance NOT because of 64 bitness by master_p · 2003-09-27 21:58 · Score: 1

Increased performance NOT because of 64 bitness

From the articles, it can be clearly understood that the increased performance of of these new processors comes not from the 64 bit data bus, but from other technologies:

1) increased number of general purpose registers from 8 to 16.

2) dedicated bus to memory for each CPU; allows for much better CPU scaling.

3) deeper pipelines

4) improved design

These things could be easily achieved with 32-bit CPUs. The fact that more than 4 GB of memory will be addressable, is a nice benefit though.

If we have not been stuck with the 80x86 instruction set, we could have seen some of these improvements earlier on the desktop...especially the additional registers.

Re:more motherboard reviews, please (esp. w/ Linux by yarbo · 2003-09-28 01:43 · Score: 1

"When I speak of free software, I'm referring to freedom, not price." - Richard M. Stallman

XP's too slow for their price...? by juhaz · 2003-09-28 02:04 · Score: 1

What the HELL are you smoking? Can I have some of that stuff, please, it's strong...

Price-performance ratio of mid-range XP's (2xxx+) still kick the living daylights of ANY other customer CPU out there.

Re:great by Wiz · 2003-09-28 08:05 · Score: 1

Well, decade = 10 years. Therefore decades has to be at least 20 years ago.

Sun were formed in Feburary 82, so by September 83 they must have had a 64-bit OS. Pretty good going!
</ULTRA_PENDANTIC>

Re:more motherboard reviews, please (esp. w/ Linux by _|()|\| · 2003-09-29 11:47 · Score: 1

FYI, I caught the following at NewsForge, linked from LWN:

SuSE Linux 9.0 Professional for AMD64 is $119.95. An update version of the software will be available online for $50.

It's not entirely clear that the update version will be available for AMD64.

Re:Finally! by Sayan · 2003-09-29 23:01 · Score: 1

Which company had the first affordable 64-bit desktop again? Sun, Apple, AMD??

Sorry all the 3 guesses are wrong. The correct answer is Digital Alpha. If you normalise the growth in technology and inflation, etc (my economics is rusty) then the Alpha 64 would win it hands down. It was 64bit when 32-bit computers were wearing diapers.

--
resurrect my .sig

Slashdot Mirror

First Round of AMD Athlon 64 Reviews In

168 of 248 comments (clear)