More Details Emerge on AMD's Hammer

Diabolus writes "Anandtech have more information on AMD's upcoming Hammer processors. " Talking with several engineers who are in the know about it, the Hammer looks pretty frickin' amazing. Itanium will have a run for its money, I suspect.

Comments

proost frooost

[Anonymous Coward]

for adequacy.org

Important Stuff:
Please try to keep posts on topic.
Try to reply to other people comments instead of starting new threads.
Read other people's messages before posting your own to avoid simply duplicating what has already been said.
Use a clear subject that describes what your message is about.
Offtopic, Inflammatory, Inappropriate, Illegal, or Offensive comments might be moderated. (You can read everything, even moderated posts, by adjusting your threshold on the User Preferences Page)

Whoa!

[Free Bird]

I just knew the Hammer was up to it, but now I'm totally convinced. x86 isn't dead yet, and IA-64 will never live!

Re:Whoa!

[Anonymous Coward]

If you're interested in more breaking developments in x86-land, check this site.

Re:Whoa!

[Anonymous Coward]

oh, so fucking funny dickweed.

it's a goatse site for everyone not interested in seeing what this perverted fuck thinks is funny.

Re:Whoa!

[Anonymous Coward]

Like I said, it's like Slashdot but more "hardware" oriented.

MHz Myth....

[jone1941]

The Itanium is going to have some competition with itself as well, comsidering the prospect of having a much "slower" (MHz wise) processor. As far as I know AMD's Hammer chips will not be slower than the Athlon XP's. This is one more area that AMD can gain a little more ground on intel.

Re:MHz Myth....

[pmz]

I really hope those on the market for Itanium machines are smart enough to look beyond MHz. On the other hand, most of them will probably be looking to install Windows, so my optimism may be unfounded.

Re:MHz Myth....

[Anonymous Coward]

Wow. Another Linux zeal^H^H^H^H user who thinks that anyone who runs Windows is dumb. It's possible that you could lose the bad attitude but my optimism may be unfounded.

Re:MHz Myth....

[Anonymous Coward]

The only MS app that will ship for IA64 (at first) is SQL Server. Hopefully DBAs are smart enough to see past Mhz scores before dropping $100K+ for a system.

Entirely different market than what AMD is shooting for, BTW.

Re:MHz Myth....

[connorbd]

I don't know -- Windows is one of several OSes competing for Itanium mindshare right now. Don't forget Linux Trillian, for example...

/Brian

Re:MHz Myth....

[ivan256]

You assume that DBA's have the checkbook. All too often, spending decisions are made by the wrong poeple.

Re:MHz Myth....

[Anonymous Coward]

What you say is true but I was responding to his implication that if a person wants to install Windows on a machine with a 64 bit processor, then they're not smart enough to understand that the clock speed is not the be-all-end-all metric of processor power. In other words he is just another /bot who likes to think that Linux users are all super smart and Windows users are dumb as rocks.

Re:MHz Myth....

[AndroidCat]

I remember when Motorola's 8 it processors took a public relations hit when they clocked at half the speed of the 8080/Z80s. Of course they did more per cycle, but Joe Buyer wants to see a big number o the chip.

Big number, heh! 2 MHz vs. 4 MHz! At this rate the "RF" noise from the chips will be visible as a deep red glow.

Re:MHz Myth....

[pmz]

Wow. Another person who thinks I must use Linux since I criticize Windows. There are more than two operating systems in this world.

I am a bit fearful

[The Slashdolt]

It will probably work well until your fan happens to go out, then it catches your house on fire. Are they selling fire extinguishers with these things? Or case smoke detectors?

Re:I am a bit fearful

[Bradee-oh!]

As TomsHardware covered a few weeks ago with a very entertaining video and supporting story, the Athlon MP/XP's have a thermal detector that can either freeze or throttle the processor when the temperature starts getting too high

As they also demonstrated, the detector does not react nearly fast enough if the heat sink is removed while the processor is operating, but it was designed for a failed HSF fan, and should manage that without a hitch.

One can only assume that the hammer processors will have as good, if not (hopefully) better thermal protection.

On a related note, it's interesting to point out that the P3 doesn't fry if you remove the hsf while operating, and the P4 doesn't even lockup causing data loss - simple throttles down until the temp is safe. At least Intel did something right...

Re:I am a bit fearful

[Anonymous Coward]

The new amd processors require motherboard support for the thermal protection to work. Yeah its different than intel but the tested motherboard at toms hardware did not support it.

What I wanna know

[UserChrisCanter4]

Has AMD unfairly optimized the processor for Quake 3?

[/sarcasm]

Re:What I wanna know

[rtaylor]

Exactly.. If they have, then I'll probably get it just for that feature.

Dual 1600 * 1200 displays would be quite nice -- so I can play Quake 3 in one and a Divx movie in the other...

Re:What I wanna know

[Anonymous Coward]

Intel is allways the one that has performed best in Quake3.. You should ask them instead! =)

Re:What I wanna know

[Anonymous Coward]

Good thing you posted that POS waste of time at +1, otherwise more people may have been able to avoid it. I may be a crap flooder, but at least I'm gracious enough to do it at 0 to give people a chance to weasel out. Oh well, it's your karma, you can lose it however you want.

Where there's smoke...

[The Slashdolt]

There is probably an AMD processor with a burnt out fan or a heatsink that slid over. Be careful with these things.

Fire

[ZaneMcAuley]

And comming soon to a home insurance contract near you.

An AMD clause "If AMD CPUs are within the perimiter of the house, you arnt insured (Act of God?) ;D

AMD's Future

[Renraku]

Why is AMD making these things so sensitive to heat? I'll bet they're also sensitive to vibration, electricity, and about anything that its competitors handle every day. Most hammers can resist hundreds of degrees before they melt/disentigrate.

Re:AMD's Future

[Anonymous Coward]

I hear there will be a military version that will cost $700.

Re:AMD's Future

[connorbd]

It's not that they're sensitive to heat per se; they just lack the safeguards Intel chips have. It's all on board on the P4, for example.

/Brian

Re:AMD's Future

[Anonymous Coward]

Something said.... not good...

Homer are you still here? You really are slow.

Re:AMD's Future

[iceburn]

I can't believe the parent was modded up as insightful. ...such an obvious joke, too.

Re:AMD's Future

[connorbd]

RTFSpec. Or go to Tom's Hardware and watch the footage yourself. Mofo sizzles like a strip of bacon. The P4, on the other hand... it's got plenty of issues, but Intel designed the thing to be damn near indestructible; it'll throttle down to 33mHz if it has to to avoid burning up.

For the record, before I got my used P2 I was in the market for an Athlon, and AMD would still be my first choice (mind you I'd check the heat sink out of the box to make sure it was in place). But I do know what the facts are.

/Brian

Re:AMD's Future

[Anonymous Coward]

Quiet Lisa, Daddy's posting on Slashdot.

Slashlies v3

[Anonymous Coward]

According to michael:

Errr, as the code is commented just above the part you pasted:

# logs attempts to break, fool, flood a particular form

That is, the people who are attempting to break, say, the comment posting form and post 500 comments at once are logged and may be banned by IP if they try hard enough.

Anyone can see, by reading Slashcode itself, not the misleading or irrelevant comments surrounding it, that the code in question does not apply to people who try to abuse the "multipost" bug (which doesn't even exist anymore; it has long since been fixed), but to people who have been moderated down with a certain frequency.

Let me spell it out for everyone since you have so little regard for the truth as to actually attempt to hide this fact:

Anyone who is moderated down four or more times within a 24-hour period will have their ENTIRE TCP/IP SUBNET banned for the following 72 hours.

That is a fact. It can be confirmed by reading the slashcode. It can further be confirmed by simply posting a comment which will be likely to garner a few negative moderations (i.e., any comment that disagrees with popluar slashdot opinions) and observing the results.

So michael, the obvious question here is, why are you lying to people? Why not just tell the thruth: The moderation system, combined with the IP-subnet-banning system, will automatically ban people for posting anything people disagree with.

Slashdot is clearly designed with the expressed purpose of surpressing unpopular opinions within its comment system.

That is, of course, your right, as this is your website and you may do with it as you choose. However, people should know that you are lying when you claim (in the FAQ and elsewhere) that people are never banned for their opinions, only for attempts at "flooding". That is complete fiction.

What to do? Well, here is a suggestion from an AC:

What we need to do now is set up some sort of IP spoofing system so we can get huge popular subnets banned from commenting on Slashdot for 72 hour periods. There must be a dozen critical subnets that would really quieten things down here if they were regularly banned. I.e. a few important University campuses.

Get going on it, trolls. Have fun with it. Let's get some of these stuffed shirts banned for a few days and see them fume.

I'm not technically adept enough to accomplish this. Anyone want to have a go at it?

Re:Slashlies v3

[Anonymous Coward]

...the fuck..?

I guess I'll never understand why some people get their rocks off trying to fuck with strangers who haven't done anything to them.

Re:Slashlies v3

[Anonymous Coward]

I think the idea is to shut down the /. community to pressure CmdrDickweed to make the system less opressive.

Re:Slashlies v3

[Anonymous Coward]

Did you even read the whole post, dipshit? Let me spell it out a little more clearly in small bites so you can understand. Here's a hypothetical situation:

1.) You make a post.
2.) Some Slashdot moderators disagree with it
3.) You get modded down (usually "Overrated")
4.) If this happens more than 4 times within 24 hours, everyone on your subnet is banned for 72 hours!
5.) When you complain about it to the powers that be (michael in this case), he blatantly lies to you about it.

Now do you get it? Yes, there are tons of crapflooders and Trolls, and those guys deserve to be banned. But there are also legitimate posters who get banned because certain people disagree with them. Then, anyone else on their subnet gets banned too.

I don't know about you, but I think that's pretty shitty.

Incase of emergency

[ZaneMcAuley]

Break case

NUMA ... shudder ....

[taniwha]

I know Linus's been talking NUMA for 2.5 - looks like there's more and more reason for it ... still historically it's been a hard nut to crack well

Re:NUMA ... shudder ....

[InsaneGeek]

Why shudder, works sweet on my Irix boxes and not too shabby on our Sequent cluster.

Sgi actually has a 64 node (128 proc) numa working on their Origin Mips line, you might want to checkout http://oss.sgi.com/projects/numa/ I think SGI is looking at the way leading this charge, and as long as SGI can stay alive long enough they'll have a good implementation. There's one thing I can say about SGI, they're scalable NUMA tech is almost beyond reproach (too bad I can't get squat for 3rd party Irix apps).

Here's the link to SGI's cat cpuinfo of their 128 proc Linux numa system running
http://oss.sgi.com/projects/LinuxScalability/dow nl oad/mips128.out

Re:NUMA ... shudder ....

[cmowire]

With respect to the Hammer, it still works in a SMP-model-like, not a NUMA model. It's not entirely SMP, but enough like SMP to make the optimizations not as hard.

Between this and Hyperthreads, new OS designs should be able to take advantage of at least multiple processors, even on the desktop. Of course, the Pentium was supposed to be the first CPU to enable SMP For The Rest Of Us, so we'll just have to see what happens.

Re:NUMA ... shudder ....

[Anonymous Coward]

Only if you interleave memory, if each processor has its own continuous chunk of the memory range then its already very much a NUMA based design. Once you start interconnecting 4/8 processor nodes in bigger configurations through HT switches then it will of course definetely become NUMA.

Its a pity they didnt put more logic (and another HT port) in to make it not need any glue for >8 configurations (hypercube for instance). But a processor for which mobo chipsets are identical for single and multiprocessor setups is cool too (is gonna drive the cost down for MP).

Re:NUMA ... shudder ....

[PeterM from Berkeley]

Hardware-wise, multi-CPU Hammers will indeed resemble NUMA. Each CPU will be directly controlling its own set of DIMMS.

However, from what I understood of the description, memory access should all be taken care of in hardware with no OS support. The CPU interconnects are supposed to make even remote memory transactions very, very fast, not much additional latency than to direct-accessed memory.

Linux would therefore "need" no explicit NUMA code, and could improve things just a bit by setting CPU affinity of a process to the one which has the process in local memory, very similar to the CPU affinity code which is already in place for keeping a process on the same processor which has its data in the CPU cache....

Maybe someone else who knows more can weigh in on this, but to me it looks like a small issue.

PeterM

Re:NUMA ... shudder ....

[Anonymous Coward]

You probably could run with no OS NUMA code, but it would/will be better with it.

First off, the core OS functionality will probably (unless AMD does something really funky) be running on the first processor to boot (the monarch). This gives the fun side effect of most system calls getting hit with extra latency (fetching the page across the cross bars - or the data from the page if you can't read-share the page itself). Plus, you have the increased contention added to the latency. The typical solution to this is to put the OS and any non-NUMA specific apps in Interleaved memory, meaning that x pages go on each processor at a time, in a round robin fashion. Spreads the load, as it were - averaging out the contention and latency... and before someone brings the point up, no you're not guaranteed that the locality is spread out (spatially or temporally) in a similar fashion, you're just playing the odds here as it were.

So, now that you've got the Global Interleave memory, you can have high-performance apps that don't _want_ to be in the interleave, or want to define their own (say the scientific research guy running particle simulation with 4 processors, coarse-grained threads -- on an 8 proc system, he doesn't want _any_ of his pages on other processors... why take the latency hit... or the programmer who knows the _exact_ memory layout and wants to define memory partitioning across processors so that each proc operates only on local memory -- can't do that if everything is interleaved). You need/want interfaces for setting up non-interleaved pools across processors or processor sets.

And that's just off the top of my head - I'm not a NUMA expert, so I'm sure I missed something.

YO!

[Anonymous Coward]

Important Stuff:
Please try to keep posts off topic.
Try to starting new threads istead of replying to other people comments.
Read other people's messages after posting your own to promote duplicating what has already been said.
Use an unclear subject that describes what your damage is.
Offtopic, Inflammatory, Inappropriate, Illegal, or Offensive comments might be interesting.

Corporate monopolies stopping progress

[euroderf]

Many may think that the Hammer is a quantum leap in microprocessor technology. Indeed, prima facie, it seems that this is the case.

But, in actual fact, this technology would have been with us many years ago were it not for corporate domination of high technology. I have been around for many years, and I remember in the late 50's and the 1960's, when computing technologies were dominated by the Universities and the public ethos was uppermost. Freedom of information reigned, and thousands of little computing groups competed to bring the new era. Unix, Multics, CP/M, Hard Drives, the Mouse, CRT displays, all these and more were made during this time.

As soon as computing became dominated by corporate interests though, a money making formula was cracked - Computer, simplistic OS, mouse, CRT, keyboard, box and sell and improve and refine. The days of real paradigm shifts stopped. The Corps controlled all, and were not interested in innovation except in as much as it forces a new upgrade cycle.

if we want computing to have a new dawn, a new time of real change, then we need to ensure that the computing industry is overthrown and controlled instead by the people. The socialist control of the means of production of hardware will allow for innovation in that realm, just as the socialist control of the means of production in software has i thanks to the GNU liscence.

I urge everyone to think of these issues, and ask themselves, do I want the same as now, but twice faster, or a genuine revolution? Constant revolution, not improvement, is the way forward.

Re:Corporate monopolies stopping progress

[Anonymous Coward]

Wow, only the first sentence of that post was actually on-topic...

The rest was clearly off-topic, and what's worse, apparently spawned another off-topic post (this one!)

Re:Corporate monopolies stopping progress

[Reality Master 101]

I remember in the late 50's and the 1960's, when computing technologies were dominated by the Universities and the public ethos was uppermost. Freedom of information reigned, and thousands of little computing groups competed to bring the new era.

What the hell are you talking about? Can you say "IBM"? That was the era of "you can have any color you want as long as its blue", unless you went with one of the seven dwarfs. Universities didn't contribute jack to anything. IBM invented just about everything during that time.

Unix, Multics, CP/M, Hard Drives, the Mouse, CRT displays, all these and more were made during this time.

...by corporations. Perhaps you've heard of AT&T (Unix, Multics)? Hard drives -- IBM. CRT -- who knows. Mouse -- this might have actually been invented at a university, I can't remember.

The socialist control of the means of production of hardware will allow for innovation in that realm, just as the socialist control of the means of production in software has i thanks to the GNU liscence.

Yeah, I know this proves it was a troll, but just in case anyone was going to believe any of that historical bullshit.

Re:Corporate monopolies stopping progress

[Anonymous Coward]

Man, this troll MUST have moderated himself up. There is no way it was legitimately moderated up as insightful. Please God, let there be NO WAY.

Re:Corporate monopolies stopping progress

[Diomedes01]

Congratulations! You just fed a troll.

Re:Corporate monopolies stopping progress

[LMCBoy]

The mouse was invented by Douglas C. Engelbart at Stanford

Re:Corporate monopolies stopping progress

[Anonymous Coward]

The mouse, like the GUI and the laser printer, was invented at Xerox PARC.

Re:Corporate monopolies stopping progress

[Anonymous Coward]

What's wrong? Ade-crap-cy too boring so you come here?

Re:Corporate monopolies stopping progress

[Anonymous Coward]

Congratulations, you just got modded to -1!

Re:Corporate monopolies stopping progress

[Anonymous Coward]

I think we all deserve some congratulations for this thread.

Re:Corporate monopolies stopping progress

[Anonymous Coward]

I'm just posting this as more encouragement for your trolling. Shit, you've pulled dozens of people in! Someday, I hope to have your trolling skills- without actually having to write a whole page of garbage, of course. It's not worth that much effort.

Re:Corporate monopolies stopping progress

[Anonymous Coward]

Who would think Hammer is any kind of leap. It's just 386 vs 286 for the modern day. AL -> AX -> EAX -> RAX??? Gimme a break. Long live IA64!

Constant revolution, not improvement, is the way forward.

This has never actually worked.

Re:Corporate monopolies stopping progress

[Reality Master 101]

Sheesh, if I was going to post an anti-technology troll, I wouldn't have made it so obvious by using words like "socialist".

Re:Corporate monopolies stopping progress

[Anonymous Coward]

I don't think it counts as a success when one troll replies to another.

wonder who will use it...

[pecka]

I just can't help but it seems that the whole industry is committed to Itanium and considering the money that allready flown into this monster and it's support the is just the OS community that can use this new processor.
Or maybe Apple will do.

Re:wonder who will use it...

[fault0]

you forget the fact that x86 already _IS_ the industry standard. x86-64 would give very fast upgrade paths to companies which have invested billions in x86 platforms.

and the industry really isn't commited to Itanium. There has been support for it, but among most circles, it's unclear who will win the next big microprocessor family round.

Re:wonder who will use it...

[Anonymous Coward]

The "whole industry" is committed to the 386, and that ain't gonna change anytime soon. If they do change, it sure as hell won't be to either Intel or AMD, because both are so weak compared to the state of the art.

Wise Intel

[eclectric]

So intel's plan to bolster their flagging market share is to introduce an entirely new platform that's not backwards compatible? I'm not exactly hardware-techno-geeky, but it seems to me this means none of your software is going to run (It would be like running a mac program on windows.) It seems to me there's got to be a way to have an intermediate step somewhere down the line that support both architectures. But then again, I could be an idiot on this point.

Either way, AMD is the smart one. extending x86 means they have 64-bit entries in the consumer market much faster, while Intel spends all of their time actually developing a new ISA. The real question will be if Intel decides to grant rights to the new ISA like they did for x86.

AMD knows, either way, that they have to wait for intel. They don't have enough of the market to be bold enough to introduce a new architecture.

Re:Wise Intel

[WasterDave]

So intel's plan to bolster their flagging market share is to introduce an entirely new platform that's not backwards compatible?

Yup, and it would have worked too (if it wasn't for you pesky kids) had the chip come out when it was supposed to. Two, maybe three years ago, with the current level of performance.

Part of what pisses me off about this whole IA-64 thing is that it was actually quite a good idea.

Dave

Re:Wise Intel

[mmontour]

Don't you think that we've hauled along the old 8086/XT baggage long enough? Do we really need a 64-bit 2GHz processor that can still run an MS-DOS 1.0 executable file, or that needs a multi-stage boot loader to crawl its way up the evolutionary ladder from 16-bit to 32-bit to 64-bit "mode", accompanied by a BIOS that has 6 different ways to map a 400G hard-drive into a 1024x16x63 parameter space?

I feel that at some point the best thing to do is walk away from the old architecture and make a fresh start with a new one. Commodore did this when they went from the C-64 to the Amiga. Users grumbled for a while, but I think that in hindsight it turned out to be the right choice - once people began to exploit the capabilities of the new platform, compatibility with the old one became irrelevant. And there's always software emulation for those cases when you really do need to preserve the old stuff.

Note that I don't actually know how much "legacy" x86 code is in the Hammer, but even the article's little picture of the register structure makes be think the answer is "too much". Anyway, when did a lack of factual knowledge ever stop someone from ranting on Slashdot? :-)

Re:Wise Intel

[RainbowSix]

Speaking of the uninformed.. let me suggest that one reason we can't switch to a new architecture is because everybody and their mother runs Windows (ignore the recursiveness of that statement!). Moving from C-64 to Amiga had people grumbling, but moving from x86 to something else is going to have people NOT BUYING. We're talking people who don't have a clue about megahertzes not mattering, new architecture is going to confuse the hell out of the normal web surfer and they likely won't buy it. These companies are of course after profit, and so have to stick with x86

Re:Wise Intel

[athlon02]

(1) What percentage of people ran (still run) DOS 1.0 exe's on a Pentium or even a 486? ... Obviously not many, so it's there as more of a "just in case" feature. (I'm not arguing with you, your point is well taken, just pointing out something that I think the x86 market *should* do eventually)

(2) Map a 400GB HD 6 ways... Oh how much I await 20GB SSDs being standard on a PC! No more mechanical HDs!!

(3) Well I'm no Amiga guru, but I am a Mac enthusiast (in addition to my Duron 750, I run a G3 500 iBook hehe)... Anyways, you'll notice that Apple made FAT binaries to support the 68k -> PPC transition, and they did the whole Cocoa thing to deal with the OS 8/9 -> OS X transition... all in all a smart move I believe.

(4) What in the world am I rambling about? Well my idea is this:

*DO NOT* get rid of x86 completely for the next few decades... *INSTEAD* you slowly fade out old legacy parts of x86 until you have a hybrid x86, where you don't have the v8086 mode, nor 16bit mode (if people run win3.x there will still be AthlonXP's for them to use, more than enough for win3.x), nor certain other outdated features. That way you remove extra hardware that *most* people don't use, cut some design costs and still have compatibility with all the latest and greatest software.

(Yes, I realize that some people do use x86 for embedded solutions, but they can easily grab a 68xx, 68k, ARM, M-Core or embedded 386/486 processor and not worry about it)

I don't mean this to sound like I'm complaining, nor that I have all the answers... in fact for anyone taking the time to read this, I'd like to hear some feedback on ups and downs people see in this.

Re:Wise Intel

[stripes]

I feel that at some point the best thing to do is walk away from the old architecture and make a fresh start with a new one.

So did DEC (later Compaq) with the Alpha. It was pretty much the fastest single CPU for floating point over most of it's life span (sometimes a new CPU would come out and beat it, but normally there would be a new Alpha within a month or two to smash it). Similarly for integer performance, but not quite as well (for example the fastest P4 systems have been beating the dead bloated corpse of the Alpha for a while in integer, but still lose out in FP). If ditching the old in favor of the new works, why are we not running Alpha machines now?

Personally I hate the x86 instruction set. I really do. I also think AMD's choice of doing the x86-64 rather then Intel's choice of doing the iTanic is a great business choice, even though it dooms us to spend another decade with the crappy 8086 compatible instruction set. Gack.

Commodore did this when they went from the C-64 to the Amiga.

I'll spare the "look where it got them" bit, and just go for...nah, just look where it got them.

Of corse as a counter point we have the Mac and it's total incompatibility with the Apple II...unless you count sharing of the ImageWriter...

Re:Wise Intel

[Waffle Iron]

It's not just 8086/XT baggage. The X86 architecture is a direct descendent of the first microprocessor, the 4004.

I for one think that its cool that we are using a vestige of the first microprocessor at 5 orders of magnitude faster speeds. It's a tribute to the human ability to create a good kludge. I wouldn't want it any other way.

Re:Wise Intel

[DahGhostfacedFiddlah]

I like the idea. Keep processors 1-generation-previous compatible. By the time the next generation comes out, emulators should be able to run the old programs at their full speed.

Re:Wise Intel

[addaon]

Don't knock the imagewriter as a major detail, though. To this day, I have an imagewriter one (the rectangular brick-of-doom, not the sleek-and-slanted iw two) connected to my mac, still working beautifully. And when I got my first mac, knowing I wouldn't have to buy a new printer was a really big deal; it was one of the things that prevented me from even looking at dos.

Re:Wise Intel

[maraist]

(1) What percentage of people ran (still run) DOS 1.0 exe's on a Pentium or even a 486? ... Obviously not many, so it's there as more of a "just in case" feature. (I'm not arguing with you, your point is well taken, just pointing out something that I think the x86 market *should* do eventually)

Probably 90% of all consumers. Ever hear of windows millenium? That new fangled OS that I don't yet need to upgrade to? It still supports all those ugly 16-bit DOS features.. Sure they did away with the DOS boot-process. But DOS is most certainly there.. And until DOS is gone (a la NT / XP), CPU manufacturers still have to support it. Never mind the fact that even Linux is based on an initial x86 boot-process. (Though obviously it's not tied to it given it's multi-platform support). But out-of-the-box x86 Linux wan't 16-bit x86 supports.

Sure win 9x is "mostly" 32 bit if not all. But it most assuredly supports the sort of legacy x86 features that both software and HARDWARE developers take advantage of.

The AH,AL 8 bit registers you see are essential to call the CPU an x86 anything. If for no other reason than IO support (don't remember the exact instructions.. it's been a while since I've read an 8086 assembly book). Note that IO is pretty much unchanged in the Athlon (since so little actually uses it anymore; relegating to windows drivers and shared memory regions).. Interrupts are also used by these 8-bit registers. In fact, pretty much anything relating to the hardware drivers (minus AGP) depends on it.

I think the loss of the ISA slots should help ease the transition.. PCI with plug and play shouldn't be too hard to port to which-ever technology superceeds. But my point is that there isn't an absence of current-market vendors that still depend on these legacy features.

Aside from hardware, x86 had lots of macro-instructions, such as using CX as a 16 bit counter, and SP, BP for string comparisons. I'm sure these are micro-op vectors in the Penium on, but they still need to be emulated and debugged somehow, thus the register set still needs to be in tact. The real question is whether they make 64bit the fast-path (requring an extra logic probagation for 32bit), or if 64bit is considered the exception.

Aside from that, I agree with you that "staging out" is the way to go. XP should help (sadly) most consumers get rid of any remaining ties to the hardware (via hardware abstraction layers; assuming that's still there). But MS has no vested interest in making the same OS for servers as for consumers. They'd love to have a win3k that only runs on expensive hardware (where they can charge a premium), with their win4Suckers running on a legacy platform that allows them to boast over 1 trillion apps served. You can't buy that sort of marketing. Heck their current strategy is to not even acknowledge that other OS's exist. When was the last time you saw an MS commercial advocate themselves over someone else. (Like AOL still has to do. "No wonder we're number 1").

Sure Linux'll support what-ever and when-ever.. That's one of it's trademarks. But 64bit has a couple down-sides (including memory / cache requirements), and having a 64bit time-stamp or file-descriptor just isn't going to impress the other 99% of the code enough to run faster - the key is going to be end-user benchmarks and or raw MHZ. That's what draws peoples attention. And people's attention is what draws MicroSoft. And as we all know.. MicroSoft rules the world. (well, it's own world at any rate).

-Michael

Re:Wise Intel

[Anonymous Coward]

The "right choice" eh? Tell me then, how come both the C64 and Amiga are dead?

Backward compatibility is awesome, and I wouldn't want to give it up. And to answer your question, yes, I do believe that its a good thing that I can still run my copy of "Ancient Art of War" on a current machine (albeit using a slowdown program). As far as I'm concerned, there are two ways that processor technology should progress..

1) Continue producing processors that are backward compatible with the x86 architecture

2) Produce an entirely new processor that is fast enough to emulate, through software, the highest end "legacy" x86 processor (ie. P4 or Athlon 2GHz).

Re:Wise Intel

[mmontour]

Never mind the fact that even Linux is based on an initial x86 boot-process. (Though obviously it's not tied to it given it's multi-platform support).

One of the nicest things about using Linux on a non-x86 platform is that you often get to use a much more advanced bootloader. E.g. on the (now defunct) StrongARM-based Netwinder, you could do a diskless boot (TFTP+NFS), specify the name of the kernel image you wanted to run (dynamically, instead of having to put it in a conf and run 'lilo'), get full serial-console support, etc. Similarly for the Mac's "Open Firmware".

The only reason x86 Linux uses the "16 bit" cruft is because it has to.

As for the Windows market, they're moving to a "subscription" model anyway in order to get a more continuous revenue stream. Once consumers are in the habit of updating all their software every (x) months whether they need it or not, it becomes easier to switch the underlying architecture. You'd use a software emulator or 'virtual machine' model to support the "legacy" software. Sure it would slow down the old apps a lot, but that's what the manufacturers want anyway so they can sell you a new chip / application with 'go faster stripes'.

Interesting points about how all the registers are used... I've never actually been brave enough to get into x86 assembly. I have a Motorola background, so I'm used to things like a flat 4G address space, "data" registers and "address" registers, and memory-mapped IO. My brain just balked at the x86 world of "memory segments", "al/ah/ax/eax", etc.

Re:Wise Intel

[Anonymous Coward]

Commodore did this when they went from the C-64 to the Amiga

Yeah, but you don't see a lot of Amiga or Commodore brand computers around anymore...

Re:Wise Intel

[mmontour]

The "right choice" eh? Tell me then, how come both the C64 and Amiga are dead?

The Amiga is dead primarily for marketing/mismanagement reasons, but also in part because it tied its OS very tightly to custom hardware. This gave it an early advantage, blowing almost everyone else out of the water in terms of graphics, sound, multitasking, etc. However it became a liability as time went on and the competing hardware improved - certain parts of the Amiga were still tightly tied to the old custom chipset. I do not believe that "inability to run C64 programs natively" was a significant factor in the ultimate demise of the Amiga.

As for why the C64 itself died, mainly just because it reached the end-point of its evolution, and the rest of the world moved on. It was an 8-bit machine, and that imposed certain fundamental limitations on it. Yes they could've clocked it up to 25 MHz, strapped on big-ass heatsinks, and added more and more bank-switched RAM, but it just wasn't worth it. Sometimes you have to walk away and start from a clean slate.

Nowadays, both the C64 and the Amiga can be emulated in software. I don't remember what capabilities the Amiga had for emulating the C64. Quite frankly, I didn't really care anymore after I had had the Amiga for a while.

Re:Wise Intel

[athlon02]

I ramble at times, so if this doesn't sound 100% coherent, don't worry, it's not you :)

Well I meant actually running DOS 1.0 exe's, not just having support for them, because even if Windows doesn't have the code in place for that, you could write a frontend or something for them anyways.

In my understanding the x86 boot process involves running BIOS code to get started, then look at the MBR for the OS's bootstrap code, and run that code, which every OS must go thru. This of course means having 16bit non-protected mode code that puts the OS in 32bit protected mode (in most OSes I would think) and from there has no ties to 16bit code and such, if the user doesn't run old legacy apps.

Oh, there's no way I want to do away with AH/AL and all the other 8bit regs, that would make work on simple char variables a bit harder and I've programmed on the Motorola 6812 so I know all about 8 bit regs and their importance. I'm talking about phasing out the hated v8086, segmentation, etc, etc... stuff eventually.

One other thing I was thinking of is if you phase out certain less used x86 features on newer processors running @ X GHz, then writing an emulator to run the code would still run with not much modification and still be the equivalent of say 200-400MHz (well beyond what a lot of old DOS and Win 3.x type apps need anyways)... So really I don't see that as too much of a problem (unlike the iTanic, er Itanium :)

OK, I'm done rambling go about your normal lives :)

Re:Wise Intel

[schwatoo]

"and they did the whole Cocoa thing to deal with the OS 8/9 -> OS X transition". Uhuh. They did the whole Carbon thing. Carbon. Repeat after me. Carbon. Not Cocoa. Carbon. Cocoa is a whole different thing (a different language for a start). Carbon == refit. Cocoa == rewrite.

Re:Wise Intel

[haggar]

Commodore did this when they went from the C-64 to the Amiga. Users grumbled for a while, but I think that in hindsight it turned out to be the right choice - once people began to exploit the capabilities of the new platform, compatibility with the old one became irrelevant.

Ummm... like, the C-64 was backwards compatible with what, exactly? Not with the VIC 20, not with the PET. And the C-16 (which came after the C-64) wasn't backwards or forward compatible, either. You could not run any VIC 20 programs on a C-64 or on a C-16. Programs created for any of those platforms were just as cross-compatible as with the Amiga, i.e. not at all!

So, what the heck was your point with Commodore?

Re:Wise Intel

[athlon02]

OOPS! You're right it is Carbon (and yes, I did know that already), and I was thinking write Carbon and not Cocoa at the time, yet I still missed that. Thanks for catching that one :)

Re:Wise Intel

[Bert64]

The AmigaOS was very extensible, you could load a new graphics.library which writes to an add-on graphics card and all OS friendly programs will run on it. there were even amiga "clones" from other companies, such as the draco, which lacked the custom hardware. The problem, was that the amiga was used primarily for games, and games were often written to directly access the hardware (for performance concerns, the os functions for graphics were terrible compared to the performance you could obtain manually.. and you could do lots of things the makers never intended) Poor marketting and lack of updates were the ultimate downfall of the Amiga tho, It had an incredibly performant OS which was way ahead of other desktop offerings from apple or microsoft, If they had got away from the image of a gaming platform, and got good application support they would still be around today.

Re:Wise Intel

[castlan]

The Commodore 128 was backwards compatible with the C-64. And at the source level , the Commodore 64 was largely backwards compatible with the Commodore PET and Vic 20, if not as fully as the C-128's "GO 64" command. (source code... :P just be careful what you POKE and where you PEEK.)

But this point mmontour was trying to make could have been better made with the transition from Apple's ][ series to the Macintosh architecture. Other than a few hardware interfaces, there was almost no backwards compatibility, and Apple planned it that way.

The Amiga was not developed by Commodore as a break from their venerable C-64, rather, the Amiga was a distinct machine from a failing company which Commodore bought, and then championed as superior to their previous offerings. Unfortunately, they just succeeded in carrying on the Amiga curse.

I never had an Amiga... I couldn't betray my Commodore 64 by dating its sexy cousin like that. Instead, I later ended up skulking around with some skanky PC I picked up at CompUSA's red light dictrict. I'm sure fond of that slinky Mac, and PCs can keep my attention by parading around in NetBSD, or some indecent Linux rags. But even in the face of a new 64 bit whore of a PC, my true love will always by my Commodore.

I dream in 8 bits.

Re:Wise Intel

[Anonymous Coward]

Of course, the ImageWriter also cost $600 (and god was I pissed that I couldn't use my Apple Scribe printer with a Mac..)

Re:Wise Intel

[aanantha]

The Itanium has hardware support for executing x86 instructions. That's about as hybrid x86 as a processor should get, I think. You can only take the x86 architecture so far before it becomes impossible to speed up no matter how much brute force Intel applies. We're getting near that point. Integer performance has been relatively decent compared to RISC processors, but the floating point is hopelessly bad. At least the integer unit is CISC. The floating point unit is even more primitive: stack based. The implementation of SIMD on x86 (MMX and SSE) is so hacked as to be almost useless. For example, MMX shares the same registers as the floating point unit. There are too few registers, and mixing SSE registers with stack based FP registers is a pain. Generating useful SIMD code is an artform on x86. The Altivec set on PowerPC is far more useful. Apple has been able to get noticeable increases in GUI speed by using it. That could never happen on x86. MMX and SSE are relegated to video drivers, games, and number crunchers. For anything else, the overhead and programming difficult outweight the benefits. And with processor speeds trailing memory year by year, you really have to start thinking about architectures that will promote efficient use of registers.

Pulling the plug on x86 may not be so hard. Microsoft is ready to provide an OS and compilers for Itanium. And all the code that people use today are written in higher level programming languages and are no longer very hardware dependant. There are already tons of old programs that will not work properly in Windows XP. People who care about running their ancient programs will not be doing so on new computers.

Hell, Apple managed to pull off the switch between x86 to PowerPC. And they did that on an OS that was written in Pascal and assembly. The PowerPC didn't even have any 680x0 emulation.

Re:Wise Intel

[haggar]

You might think I am a stubborn bastard (that's your constitutional right ;-)) but his point with Commodore was totally off hte mark: before the Amiga, Commodore didn't give a flying foot about backwards compatibility. Actually, even the hardware was MADE incompatible: you certainly remember how the C-16/C-4plus/C-116 had completely different connectors for the peripherals, than C-64. And as I said, no code compatibility between PET, VIC-20, C-64, C-16/C-4Plus. It's actually with the Amga that Commodore started thinking about backwards compatibility: you had the Amiga line of computers that was evolving something like this: Amiga 500, Amiga 2000, Amiga 600, Amiga 1200. Did I miss any?

But I agree with you that his point would have much better made with Apple ][ vs Mac.

BTW, I wish I had an Apple ][. Emulators don't cut it ;-)

Re:Wise Intel

[Ziviyr]

Moving from C-64 to Amiga had people grumbling

Sorry, it was awfully painless. Amiga was a whole new architecture. Few people were grumbling about leaving 160x200 4 color graphics behind for 320x400 psudo-4096 color graphics. Heck, it blew 8 bit out of the water, 32-bit cpu! Actual wave-table synthesis, gave rise to the .mod scene. SID tunes are painful to hear most of the time.

Oh, and Amigas started to have harddrives put on them.

Saying Commodore 64 users grumbled when they moved to the Amiga is like saying residents of hell grumbled when Mr. Snowcone made his rounds.

Re:Wise Intel

[rhost89]

Well as much as i would love to get rid of the 8086 baggage, I think they are alternatives that would be just as easy that would allow us to keep the legacy. As an asm programmer all that i would want is 1 opcode and some acompanying microcode for each mode of the processor. For instance init32 would do all the work i would have had to do myself of creating my GDT, LDT an IDT (Global, Local, and Interupt descriptor tables) initiate memory paging and other asorted stuff. It wouldnt, actually i should say couldnt, fill out any of these tables for me, but it would be a start. Another feature i would like is the ability for all the controlers on the system bus (IDE, Video, Sound, etc..) to have an interface that would be easier to work with in asm. Eg. access the HD without LCHS->LBA->PCHS, or init a gfx mode, and be able to write to a frame buffer more easily. (VBE was a nice thought, but poorly implemented, and is somewhat dificult to use.) It would just as easy to do this for 64bit mode. Actually i dont know why intel has not done this before, i would save me quite a bit of code to bootstrap and init an OS.

Re:Wise Intel

[vidarh]

Actually, you can buy a 20MHz 16 bit drop in replacement for the old 6510 CPU in the C64, and you can also use up to 16MB of RAM with it, 1 GB HD, etc. Take a look here, and be very, very afraid :-)

In other words: Some diehard fans actually found it worth it...

While the C64 and Amiga scenes may be mere shadows of what they were in the past, they still exists.

Re:Wise Intel

[Mr Z]

Interesting points about how all the registers are used... I've never actually been brave enough to get into x86 assembly. I have a Motorola background, so I'm used to things like a flat 4G address space, "data" registers and "address" registers, and memory-mapped IO. My brain just balked at the x86 world of "memory segments", "al/ah/ax/eax", etc.

It's really not that hard. AH and AL map to AX on x86 the same way the A and B accumulators on the 6800 map to the D accumulator. (Ok, you really meant you prefer 68k, not all things Motorola.) Ignoring the 8-bit sub-registers, you have 7 16-bit general purpose registers to work with -- AX, BX, CX, DX, SI, DI, BP. (Ok, someone will scream "BP isn't a general purpose register." Just turn on -fomit-frame-pointer when calling GCC and get on with life. Or go play in traffic.) These registers can hold data OR addresses -- no partitioning between the two uses. (I just heard someone squeal about addressing modes, and maybe something about string instructions... Most of the time, it's not an issue. Hey, aren't you the same guy who was squeaking about BP?)

So, anyway, how is that harder than having to deal with separate data and address registers? (Why partition registers by functionality?) And as long as all your data fits in 64K, you're all set--you never have to think about segments. ;-) (Hey you, with the GBs of porn... put that thing back in your pants. And stop listening to those MBs of pirated MP3s.)

Now the 32-bit extensions are even easier, if you ignore the 16-bit ways of referring to registers (most of the time, it's better that way anyway), you just put an 'E' in front of the name and they're all 32 bits. Not too hard to think about. And at least under Linux (I claim no knowledge of Windows), you get a nice 32-bit address space.

So, see, it's not so bad. (Hey you, snickering in the back! Cut it out.) The x86-64 extensions just continue this tradition -- as long as you ignore the old crap, the new crap isn't all that horrible.

All that said, I agree that x86 is a pretty grotty CPU architecture. :-)

Exercise for the reader: Identify uses of sarcasm and ironic humor in the above post.

--Joe

Re:Wise Intel

[QuickFox]

Amiga 500, Amiga 2000, Amiga 600, Amiga 1200. Did I miss any?

Yes, before that there was the Amiga 1000 (yes, the 1000 did come before the 500, very curious).

I bought one of the ten very first Amigas sold in my country, Sweden. I waited at the shop's doorstep as they came in, more or less. It was the incredible Amiga 1000, miraculous technology in those days, light-years ahead of anything else. Aaahhh, those days... Sniff...

It was a strange feeling to go to work and sit down at the stone-age "professional technology" PC.

Give a man a fish, and he can eat for one day. Teach him to fish for you, and you can eat for a lifetime.

Re:Wise Intel

[Anonymous Coward]

It's probable a lite late to ask this but any one knows If the new windowd xp 64bit supports some type of emutlation for the standard win32 if that is true then probabla tat will solve the problem of runing old applications on the itanium platform.

Re:Wise Intel

[Anonymous Coward]

Just to point out, I didn't have any problems running my VIC-20 programs on my friend's C64. My cassette tape drive was compatible, and the tapes loaded fine. Also I seem to remember having at least two programs on cartridges (one was a game called "Aliens" or something and the other was a Solar System educational cart) which both worked on his C64.

Hell, my VIC-20 joystick even worked with my old Atari 2600.

Re:Wise Intel

[Jage]

As a long time Amiga user (1989-1994), I can assure you Amiga *doesn't* have wavetable synthesis. Wavetable doesn't mean anything like "a whole bunch of samples", but it is a *synthesis* method. Just like FM-synthesis, waveguide, etc.

An Amiga could merely play 4 sound samples simultaneously. Two channels were assigned to left channel and two to right channel. Each channel also supported 6-bit volume (64 levels).

What, 68000 a 32-bit CPU? Since when? A 68000 can address 24-bit address space and it has 16-bit data bus. Sure, internally it is a 32-bit CPU, but using 32-bit ops took longer than their 16-bit counterparts! Smells badly as if the ALU was just a 16-bit one... (don't know this for certain).

Re:Wise Intel

[maraist]

(Why partition registers by functionality?)

Well, simply it makes for faster CPUs. (unless most of the time you're physically interchanging from GPR to Address Registers). The Digital Alpha, for example, went even further and utilized two completely separate register sets for GPRs. I don't remember if the programmer was required to not perform operations that would pull from both register sets or not (e.g. was it just a caching localization, or was the bottom half and top half of the register address space physically separate).

The main advantage is a minimazation of ports on the register set and a reduction in the number of buses. Each execution unit typically requires one write port for each register. If you have 6 integer execution units, then that's 6 write ports (and probabably something like 6 read ports, but in theory 12 read ports). Each port requires an address decoder and extra levels of probagation in the register fetch stage.

Back in the old days, where we didnt see heavy pipelining (especially in first generation 68K), this was expensive and slow. The 68K was clean in many ways, which included separation of dissimilar functionality to segregated addresses and buses (and possibly execution units). Since there's no contention between addressing and general ALU operation, it's closer to true divide and conquor. Mix in the fact that the 68K CISC core could utilize op [Mem] = [Mem], [Mem], the load on address registers and logic was pretty heavy (at least in comparison to RISC architectures).

I once did a simple CPU design project which unified the FP regs and the int regs. The focus was on interchangability of data-types, and simplicy of design.. But what I quickly found was that in almost all cases (except register exchange) things were worse off.. The large register set had to have exteraneous fields to handle the various datatypes (even if they weren't used 99% of the time). That logic took extra probagation layers. Additionally, the number of address bits in the assembly code was upped (since fp ops couldn't assume a separate address space than int ops). Plus I found that the number of ports I had was horrendous.

Arguably, address calculation more regularly requires utilization of integer units, and thus there will be a significantly higher percentage of swapping betwen GRP and AR than between FPR and GRP. None the less, Motorola found it advantageous to do it that way.
Once Load/Store become popular (as with the PowerPC), the benifits of separate addressing fell off. (Number of mem accesses / instruction was now well below one).

Re:Wise Intel

[Ziviyr]

Whelp, my bad, I thought it was a cool name for what the Amiga had.

Those 4 channels made up a pretty cool MOD standard.

If the core is 32 bit then its 32 bit. '020 and up were full 32 bitters also.

Re:Wise Intel

[haggar]

Well, I guess you are talking about some BASIC programs for the VIC 20 you wrote, which didn't use any of C64's hardware capabilities? I remember, even the number of columns of thext was different, not to talk about resolution, sound, mapping, memory, different CPU even.... From that point of view, your VIC 20 programs might have worked with most BASIC interpreters.

Re:Wise Intel

[Jage]

Well, the instruction timings oughta be the same for 16-bit and 32-bit instructions, which wasn't the case with 68000 (except, of course for multiplication, division and other instructions which actually have more work to do). But, then, 68000 only had 16-bit multiplication and division anyways... (32-bit/16-bit and 16bit*16bit). Even that should say something.

Mhz to the stars because of the pipeline?

[Andreas(R)]

"very deep pipelining (20+ stages) to attain very high clock speeds (10GHz+ by 2006)"

Did AMD design the pipeline to get very high clock speeds just for marketing purposes?
(because everyone they realised performance rating is doomed?)

Still, a very promising CPU. (but a lot can happen before the CPU is realeased sometime in 2003)

Good luck AMD!

Re:Mhz to the stars because of the pipeline?

[Glock27]

Did AMD design the pipeline to get very high clock speeds just for marketing purposes? (because everyone they realised performance rating is doomed?)

No, like Itanium Hammer gets more done in a single clock. I can't see how this will hurt Hammer vs. Itanium. Itanium/McKinley will have much lower clock numbers than Hammer, which is supposed to debut at 2+ GHz.

I also disagree regarding the performance rating being 'doomed'. I think the jury is still out, and so far it has had the effect of raising ASPs (average selling prices) for AMD processors relative to the overrated P4 - which is a good thing for AMD.

Still, a very promising CPU. (but a lot can happen before the CPU is realeased sometime in 2003)

A small correction - Hammer is supposed to ship in small quantities in the first half of 2002, and in production quantities in the second half. The processor roadmap appears to missing at the moment, but the release date is mentioned in the x86-64 FAQ.

I can't wait to see the first SPEC numbers!

299,792,458 m/s...not just a good idea, its the law!

Re:Mhz to the stars because of the pipeline?

[Anonymous Coward]

> "very deep pipelining (20+ stages) to attain
> very high clock speeds (10GHz+ by 2006)"
>
> Did AMD design the pipeline to get very high
> clock speeds just for marketing purposes?
> (because everyone they realised performance
> rating is doomed?)

Your comment about "high clock speeds just for marketing purposes" is especially interesting since your quote "to attain very high clock speeds (10GHz+ by 2006)" was _not_ about AMD, but Intel!! Here's your quote in context:

"It's easy to assume that AMD's next-generation CPU has to follow the same basic principles as Intel's NetBurst architecture:

1) very deep pipelining (20+ stages) to attain very high clock speeds (10GHz+ by 2006)
2) small, low latency caches and a very high bandwidth memory subsystem (e.g. RDRAM)

However it would make little sense for AMD to completely depart from their K7 architecture in order to make their future CPUs more P4-like."

Backwards compatability big advantage

[Zergwyn]

It looks like the future of CPUs is definitely 64bit+. The Itanium, Hammer, and G5 are all 64bit processors. However, it will be a long time before a lot of applications are rewritten to take advantage of 64bit architectures. In addition, some applications won't actually benefit at all, and are therefore unlikely to be recoded for quite a while. Therefore, how each of these processors runs legacy code is important.

From the look of it, both the Hammer and the G5 can run old, 32bit code natively. This means that today's apps will continue to be able to run at top speed on the new chips, because the instructions still exist in hardware. This is definitely good for people with lots of older apps(ie, almost all of us.) However, a lot of the reports on the Itanium seem to indicate that, in making a completely clean break, it is forced to emulate older 32bit instructions, resulting in an actual -slowdown- for many programs. Eventually, Intel's clean break might give it some advantage, and that advantage might come quickly for the big metal server market. However, it seems that AMD will be able to win out on the desktop. Of course, here we are comparing rumors on a rumored chip to a different unreleased chip, only Bob knows exactly what will happen between now and release time...

Re:Backwards compatability big advantage

[yesthatguy]

To some degree, it will merely be important to have good compilers written for the 64 bit processors. If programs are written in a fairly popular high-level language, a simple recompile ought to bring a decent amount of optimization for 64 bit processors.

Re:Backwards compatability big advantage

[jmauro]

Itanium can run un-modified x86 and in certain cases PA-RISC binaries unmodified. Look at the specs, there was no clean break. Intel learned with the i960 and the 8080 that clean breaks are not liked by those designing the systems at all. The x86 stayed around and will continue to stay around for as long as Intel stays around. Intel will have nothing else.

Re:Backwards compatability big advantage

[NutscrapeSucks]

However, it seems that AMD will be able to win out on the desktop

OK -- where's the software support? Where's Windows/AMD64? Where's the need for 64-bit desktop chips?

I like AMD's strategy in theory, however it will be marketed like a box of Cheerios that says "NEW - Now With More Bits!!" and nothing really to back it up.

(I should note that Apple has a similar problem with the G5, except they will ship native OS support, and it's concievable that a 64-bit CPU will have an advantage for media applications, which is pretty much their only market. Will 64-bit Quake or 64-bit OpenGL drivers help that much?)

Re:Backwards compatability big advantage

[roca]

> Itanium can run un-modified x86

Slower than a 150MHz Pentium, yeah.

Re:Backwards compatability big advantage

[whovian]

The bit about rewriting applications was also said when Apple transitioned to the Powermac from 680x0. It took many many months (I estimate up to 2 years? -help!) for various softwares - including MacOS - to catch up to being native code. This full backwards compatibility of Hammer and G5 from the start deserves kudos, in part because emulators will not be needed.

Re:Backwards compatability big advantage

[Anonymous Coward]

The big iron is where the money is. Intel has the means to compete at the low end for quite some time, while increasing market share at the top. AMD will not get into the back office like Intel and will be consigned to making its living off of smaller and smaller margins.

Re:Backwards compatability big advantage

[Anonymous Coward]

oh boy, you mean i'll be able to run my, oh, ZERO low-level HP-SUX apps natively on an Itanic? SIGN ME UP!!!

honestly people, hardware compatibility is not that big of an issue. witness alpha. emulated x86 just fine in software, rather quickly too, probably because the chip was able to be kept smaller and more streamlined by keeping the emulation out.

Re:Backwards compatability big advantage

[tandr]

man, I am software guy, but when I read i960 spec, I was crying out loud for poor silly Intel -- "why the heck they are stuck with x86 if they have THIS ?????".

BTW, when Apple wanted new processor, they took i960 very seriously. You know what Intel answered them ? "We are not interested in this product". So, PowerPC won.

ah... too late anyway.

Re:Backwards compatability big advantage

[The Man]

It looks like the future of CPUs is definitely 64bit+

No, the present is 64bit+. The peecee is the only type of workstation or server still shipping with 32-bit CPUs. Sun killed their last 32-bit workstation in 1998. Alpha's been 64-bit forever. SGI has shipped 64bit CPUs since the Indy/Indigo2 and has been running 64bit IRIX by default on everything since last year (the holdout? O2, interestingly enough, because of bugs). I could go on... The reality is that you can already buy 64bit workstations running 64bit OSs with good performance for less than $1000, and in some cases only a few hundred.

The reality is that the peecee is way behind the times.

Therefore, how each of these processors runs legacy code is important.

Very true. Unfortunately neither is really getting it right. For examples of how to support mixed 32bit and 64bit binaries and even OSs on the one 64bit CPU, see the MIPS3 documentation. For a cleaner transition that required changes at the OS level only, take a look at the SPARC V8 -> V9. Can you say "seamless?" I knew you could.

The trick, naturally, is to design a proper instruction set to begin with. Then you can extend and enhance it easily without having to break backward compatibility. Too bad Intel didn't realize that.

Re:Backwards compatability big advantage

[maraist]

honestly people, hardware compatibility is not that big of an issue. witness alpha. emulated x86 just fine in software, rather quickly too, probably because the chip was able to be kept smaller and more streamlined by keeping the emulation out.

Initially, the ALPHA's speed was due to "leaving complexities out"; it's minimalist approach to assembly (including a fuzzy FPU which was very fast if you didn't need ieee precision). But it definately didn't leave the emulation out cold. If I'm not mistaken, the ALPHA had a huge side ROMish type of thing that allowed VAX complex instruction translation lookups.

Fx86 worked fast because it incrementally translated x86 to native ALPHA. Drivers and OS libraries were already native. Thus only a moderate fraction of your code ever ran under emulation (given a long enough lifetime).

The reason hardware compatability is an issue is that if you don't have the R&D to port to multiple platforms, you choose the one that'll make the most money.. It's rarely your problem that things run too slowly; especially if the uppitiest customer will be willing to shell out for a maxed out current-state-of-the-art x86. (with proprietary motherboards that use faster memory, etc).

But even the ALPHA has legacy problems, as they're violating their "minamlist" approach by introducing out of order execution in their latest processors...

Oh how Alpha is missed.. I cheered the K7 because if I saved up enough money, I could get the Alpha version

-Michael

Re:Backwards compatability big advantage

[fault0]

Either way, it will be the next generation x86 processor as well, wether or not people use the 64 bit stuff (which people will, if they sell it to a broad base market).

Re:Backwards compatability big advantage

[connorbd]

Besides, Intel owns StrongARM now anyway -- presumably this happened long after Apple went hunting (probably 1991 or so, but I don't remember for certain), but they've got a strong RISC platform if they want to push it. Obviously they don't care to, so...

/Brian

Re:Backwards compatability big advantage

[stripes]

The trick, naturally, is to design a proper instruction set to begin with. Then you can extend and enhance it easily without having to break backward compatibility. Too bad Intel didn't realize that.

The SPARC and many other RISCs had a "seamless" 32 -> 64 bit transition mostly by doing two things.

1. They added 64bit load and 64 bit store instructions (existing load and store remained 32 bits). All the other stuff (register to register instructions) went to 64 bits.
2. Made large (incompatible!) changes to the supervisor mode. This only matters to the OS and boot loader, and Sun owned the dominant OS on the SPARC boxes, SGI owned the dominant OS on the MIPS boxes, and they made all the changes to the OS as needed.

There is no reason Intel/AMD couldn't make new 64 bit load and store instructions, and redefine all references to EBX (and the other 3 registers) to be 64 bits. That would work just fine.

The part that would suck is Intel and AMD do not own the OS, or even the bootloaders that runs on their CPUs! MS, and a handful of BIOS makers do. They would have to be convinced it is worth it to do anything.

NOTE: I'm not saying the x86 instruction set is anything close to well designed. It is a shambling horror, but extending it to 64 bits is not really harder then extending the SPARC to 64 bits. In fact if you look at what AMD did it is a pretty easy change (and I think the article is wrong, you can use the new 4 GPRs without having to do any 64 bit stuff, but the OS still needs to be changed to save and load the extra registers).

Intel merely decided the 32 bit to 64 bit change seemed like a good time to try to make a play for the high end market, and to do that with a new instruction set. That might have even been a good idea if they hadn't screwed it up enough that the itanimum earned the nickname the itanic...

Re:Backwards compatability big advantage

[Potlucker]

Surely you're not suggesting Microsoft Bob is joining forces with Clippy to help predict the future?!

The end is truly near.

Re:Backwards compatability big advantage

[ppetrakis]

Oh yeah, backwards compatbility, Have you looked at the SPEC scores for their compatibilty? About what a Pentium 90MHz can do... To this very day Alphas can emulate the x86 instruction set FASTER in SOFTWARE than intel can in hardware. FYI that would be equivilent to atleast a PII 400 on an EV6 alpha machine.

As for porting applications you don't even have to do that right away. The DEC/Compaq compilers have the abillity the only use the lower 32 bits of the address space. Which is sometimes the only way to get an application for you platform like netscape. BTW, That compiler team works for INTEL
now.

Also the integration of LDT aka hypertransport makes hammer capable of some extreme throughput.
You can learn more about hypertransport at http://www.api-networks.com .

As for microsoft. It's really hard to justify the development and ongoing costs of developing XP and the entire backoffice lineup when the platform it runs on is barley into production and has zero market presence or readily available killer apps.

Draw your own conclusions but make no mistake. No one is out of this game yet. That includes AMD, Intel, Compaq, AND HP.

Peter

Re:Backwards compatability big advantage

[jmauro]

Itanium can run un-modified x86

Slower than a 150MHz Pentium, yeah.

I said it could do it, I didn't say it could do it well.

Re:Backwards compatability big advantage

[tonywong]

Unlike the x86 ISA, the PowerPC architecture has already been defined (since PowerPC's inception) with 64-bit goodness, it just hasn't been implemented on the chips yet.

Re:Backwards compatability big advantage

[SEE]

Itanium can run un-modified x86

Yes. Amazingly, though, it runs x86 slower than a software-emulation package on competitors' RISC chips.

Re:Backwards compatability big advantage

[Bert64]

So, why did intel develop a new 64bit architecture, which requires new applications to take advantage of it, and runs old applications under very un-performant emulation. They could have extended the Alpha or maybe the PA-RISC architecture, Both of which are 64bit and have been for some time, and already have reasonable levels of application support, and atleast in the case of the alpha, a relatively performant x86 emulation.

Re:Backwards compatability big advantage

[morbid]

"Itanium can run un-modified x86 and in certain cases PA-RISC binaries unmodified."
Yes, but only by using a software emulator. The hardware has no support for these instruction sets at all. This is just like the FX!32 x86 emulation software for the Alpha. It suffers from a fair old overhead and performance decreease, not lesst because in the case of Itanium you are trying to map a program written for a (essentially) serail processor onto a VLIW processor. Not easy.

Re:Backwards compatability big advantage

[Tower]

>Surely you're not suggesting Microsoft Bob is joining forces with Clippy to help predict the future?!

"Hey Skipper! Wouldn't that just be the most wonderful thing? We could help write those pesky memos for you, and leave you plenty of time for more help to play Microsoft Golf, Skipper!"

Re:Backwards compatability big advantage

[Anonymous Coward]

No, you can't use the new registers or even 64 bit
instructions in 32 bit mode (64 bit reuses the one
byte inc/dec instruction opcodes as prefixes).

Re:Backwards compatability big advantage

[jmauro]

From what I understand of the processor design this isn't true at all. It's not software emulation running on the OS like FX!32, but embedded in the hardware its self. It can boot an IA32 os like Dos 6.22 or Linux or Windows XP. The IA64 instruction set can detect and identify IA32 instructions and execute them in seperate hardware. Some instructions are treated in the way you descibe (but even then its in software on the chip), but the vast majority are treated in the hardware its self. The reason its so slow is that the IA64 wants three instructions per clock cycle. IA32 is really one instruction about ever other cycle. Read here for more information.

Integrated Northbridge

[Cytotoxic]

There is a lot here to stimulate discussion. Most obvious is the decision to continue with x86 backward compatibility, but more interesting to me is the integrated memory controller. This should lead to much higher memory access performance, but I'm a little worried that it will lock the processor into a particular memory architecture. Alternately, it could allow motherboards to accept newer memory types with a CPU upgrade, since there is no northbridge on the motherboard anymore. Not being a board architect, I don't know for sure what the implication is for future upgrade paths, but the integrated northbridge is definitely a bold step that should help aleviate one of the biggest bottlenecks for the CPU.

Re:Integrated Northbridge

[cmowire]

I have considered this myself.

With regards to locking a processor into a particular memory architecture, that shouldn't be a huge issue. For one, most processor architectures stay with the same memory architecture in the chipsets for a useful span of time. So a non-issue that way, IMHO.

Now, about changing CPUs and getting a better memory architecture, that's not extremely likely. A newer memory architecture will probably have different shielding/terminating/etc. requirements. The l33t motherboard manufacturers will probably make their boards have enough headroom that it might be able to take the new memory architectures.

But that's virtually impossible to work. If it works on my buddy's p1mp ASUS motherboard, so if I have a cheappie bargan-basement motherboard, I'll expect it to work. Except that the cheappie motherboard wasn't designed with headroom.

AMD nets one happy customer and one very pissed off customer. So they will probably change things or put configuration pins in there so that the first crop of DDR333 motherboards will do a maximum of DDR333, no matter what.

Plus, most rational people upgrade processor and motherboard at the same time anyways.

So it's probably a non-issue. I personally think the integrated northbridge has been a good idea for a while. I want a 4 or 8 CPU Hammer. ;)

Re:Integrated Northbridge

[Junta]

As to the comment about allowing memory architecture without CPU change, it's highly unlikely. Even with the northbridge on the CPU, different memory architectures are rarely ever pin-compatible. So, in essence, both CPU and motherboard absolutely must match in terms of memory, whereas before you could theoretically have either AMD or Intel with, say RDRAM or SDRAM (though RDRAM is really stupid..)

Re:Integrated Northbridge

[Nelson]

It's a killer idea to integrate it for AMD. To finish the job they should also integrate a fair amount of the south bridge as well. That was one of the bigger bumps with the athalons initially, they had a great chip but no south bridge to drive it and then there have been flakey ones.

AMD has been battered but they can still run with intel on cost, so long as motherboard makers can buy support chips dirt cheap and that's where intel can really take off.

Skip Itanium

[MlBruehlly]

You bet Itanium will have a run for its money, thats why I'd wait and buy a McKinley. Itanium is only a development platform, McKinley is packing more muscle than Itanium ever will...

Re:Skip Itanium

[Anonymous Coward]

Yes, of course, we've been hearing that for years. Forget Merced, McKinley (more vapor-than-vaporware) will be much faster -- just like Windows will be stable someday. People are way too willing to settle for mediocrity for promises of excellence rather than taking the better solution from the littler guy.

ISA bus

[Alien54]

There is some criticism of the Hammer chip in the support of ISA. There is a more basic problem because limitations of what things you might want to add on, vs what you want integrated in the motherboard. The problem of what to do with the old ISA bus is mostly an issue of the old installed base. But it is still useful for some basic cards.

The limited number of PCI slots (on home systems) vs ISA slots makes it an issue for people who want to have a system like this

1. PCI SCSI
2. PCI Modem
3. PCI Firewire
4. PCI IDE Accelerator
5. PCI NIC
6. PCI Sound Card
7. etc

I presume the video is AGP.

Yes I know people who would do things like that. Ultimately this one guy will have his capabilities spread over two systems, because he cannot fit it all into one, not with a major balancing act.

Re:ISA bus

[cymen]

Or he could buy a good motherboard with decent onboard IDE, NIC, Sound, even Modem (or get an external modem), and Firewire. Then all he has to do is put in a SCSI board. Even that can be integrated on an expensive board...

Sorry but your example only holds water for people stuck in the stone age of motherboards. Some motherboards have good integrated peripherals. People who want everything on a card can buy two or three systems as far as I'm concerned. Who cares about the few nimrods who want to do this?

Re:ISA bus

[Screaming Lunatic]

ISA is for suckers. Nobody releases ISA products anymore anyway. You have to remember that Hammer won't be prevalent in the market for a couple of years. ISA will have been dead for a while. PCI is also for suckers. The networking will move onto USB etc. Which makes sense, you want to be able to connect networked products, such as your Palm, on the fly. Just like AGP is dedicated for graphics, don't be surprised to see something dedicated for SCSI as it becomes more prevalent. You don't need IDE if you have SCSI. There you go, no more concern over limited number of PCI slots. SL

Re:ISA bus

[Alien54]

well, some folks you work with are never happy. They will want to put everything into it.

These are often the same guys who want to return CPUs after three months because the price fell.

Re:ISA bus

[toupsie]

The limited number of PCI slots (on home systems) vs ISA slots makes it an issue for people who want to have a system like this

1. PCI SCSI
2. PCI Modem
3. PCI Firewire
4. PCI IDE Accelerator
5. PCI NIC
6. PCI Sound Card
7. etc
I presume the video is AGP.

Gee I forgot what it meant not to own an Apple PowerMac. All those items you mentioned are stock on my Dual G4/500 motherboard excluding my Adaptec SCSI PCI Card. I feel for you man, I would have to be saddled with ISA slots. What a waste.

Re:ISA bus

[Anonymous Coward]

network, sound and IDE are all part of any recent motherboard. That still leaves you with 5-6 PCI slots and an AGP slot to play with. Modem's should be attached via USB or serial ports, not internal.

no system should need both scsi and firewire.

Re:ISA bus

[RedWizzard]

You're confused. In this context ISA means Instruction Set Architecture not ISA bus. It is the job of an IO controller chip (traditionally the South Bridge) to provide IO buses. The CPU has nothing to do with it unless it's an embedded or system-on-a-chip type of thing.

Re:ISA bus

[Anonymous Coward]

PCI is for suckers? What the hell?

Expect to see faster, 64-bit PCI. But don't expect it to go away. Not by any means.

Re:ISA bus

[Anonymous Coward]

What a strange thought. You want to be stuck with the crap that comes stock on your mobo? And how would you upgrade it? Toss the whole system?

Re:ISA bus

[haruharaharu]

Perhaps you could tell me how much i can upgrade a decent 10/100 NIC or a ATA-100 ide controller.

Re:ISA bus

[cymen]

Oh, that is simple! You buy a new motherboard. Seriously.

Ok, so in some cases it doesn't make sense but lets be realistic - if you have to upgrade everything on the motherboard you can easily afford to simply buy a new one than spend more on individual parts. If you need to only upgrade a few simple things like IDE or a NIC than go ahead. What's holding you back?

Re:ISA bus

[cymen]

What a strange thought. You want to be stuck with the crap that comes stock on your mobo? And how would you upgrade it? Toss the whole system?

Look buddy - get your head out of the days of yore with emachines and such fruit filled computers. Go take a look at new motherboards. The day of bargin bin intergrated motherboards is passing. Sure you'll still find some crap among the perls but you will find perls these days. Amazing, isn't it? But don't take my word for it - go look.

Re:ISA bus

[haruharaharu]

You missed the question - how much can these components be upgraded? ATA-100 is pretty much the top for most systems (unless you get a board with ide raid) and a decent 10/100 net card is as much as you're likely to ever need. My point is that mature stuff goes on the motherboard.

Re:ISA bus

[cymen]

Sure, ok... What is your point? Someone is just going to come along and bitch about the lack of ATA-133 support or 1 Gb ethernet. It's a moving target, not a fixed target. I agree completely with you but I don't see why you are making the comment.

Re:ISA bus

[don.g]

ISA is *simple*. If you woke up one morning and thought "hmm, I'd like to build an interface card for $SOMETHING", and you wanted to produce it in a quantity of 1, then you'd probably want to build an ISA board.

Re:ISA bus

[Bert64]

USB was never designed for networking, and currently cannot compete with a high end ethernet card. Plus, i can connect devices to my ethernet port on the fly. And you could always have multi function cards, most of the SUN SCSI cards have integrated ethernet for instance.

And NetBSD already runs on it

[jacexpo069]

Even before the processor is out, NetBSD already runs on it. See here

Re:And NetBSD already runs on it

[markhahn]

Linux has been running in it for some time, quite possibly beating *BSD. good talk on this at OLS.

Re:And NetBSD already runs on it

[Anonymous Coward]

yeah, except that linux is for bitches.

Just Like You.

Re:And NetBSD already runs on it

[o_kenway]

Yayyyy!!

I don't even have one yet and I won't have to run linux on it when windows fails to run.

Good old NetBSD!

enough x86/CISC already

[Anonymous Coward]

Give me Alpha or give me death!

Sorry AMD/Intel, you can suck my cock. Your processors are not a technological bite in the ankle of Alpha, nor do they sing with the elegance of ARM.

Since Sun/SPARC is year after year of "nice quotes, must try harder", I feel it's time for an FHF. Or is there hope in PowerPC?

This is good news indeed

[Anton Anatopopov]

I wonder if they will do to the Itanium what the Athlon did to the Pentium and the Duron did to the Celeron. I can see a real market for cheap 64-bit processors. Lets hope they get their pricing correct ?

If I had a Hammer...

[Anonymous Coward]

With apologies to Arlo Guthrie...

If I had a Hammer
I'd troll in the morning.
I'd troll in the evening.
All over slashdot!
I'd troll out goatse.cx!
I'd troll out porno!
I'd troll out love between
sllort and perdida
All over slashdot!

If I had Itanium
I'd troll in the morning.
I'd troll in the evening.
All over slashdot!
I'd troll out BSD!
I'd troll out "dying!"
I'd troll out fights between
michael and crapflooders
All over slashdot!

If I had an UltraSPARC
I'd troll in the morning.
I'd troll in the evening.
All over slashdot!
I'd troll out Katz flames!
I'd troll out first posts!
I'd troll out feuds between
trolltalk and adequacy
All over slashdot!

Well I've got a Hammer
And I've got Itanium
And I've got an UltraSPARC
All over slashdot!
It's the Hammer of crapfloods!
It's the Itanium of flamebait!
It's the SPARC of the struggle between
the Editors and the Trolls...
All over slashdot!

Hammer will rock!

[Glock27]

Linux has already been ported to the simulator, and supports 511 GB of memory per process. That should do for a start!

Each feature of the Hammer taken alone is evolutionary, but the overall effect should be revolutionary (at least with regard to Intel server market share;).

AMD stock is looking like quite a bargain at around $10/share... :-)

299,792,458 m/s...not just a good idea, its the law!

Re:Hammer will rock!

[jacexpo069]

However, NetBSD was the first OS on it. See here for the Wasabi press release

Re:Hammer will rock!

[Anonymous Coward]

> and supports 511 GB of memory per process

One slimline kernel. 800k.

One vi session. 200k.

One gcc compile. 9000k.

One demand-loaded shared glibc. 3000k.

511.99 GB free virtual memory for a Windows XP install under VMware. Priceless.

Re:Hammer will rock!

[guru_steve]

Linux has already been ported to the simulator, and supports 511 GB of memory per process. That should do for a start!

It's too tempting not to insert a joke about Mozilla here...

Re:Hammer will rock!

[Anonymous Coward]

get OVER yourself. NetBSD is good. Linux is good. They both run *Hammer. Go to hell and DIE. (in that order). Quit trying to raise trouble all over this story.

Re:Hammer will rock!

[fault0]

one correction:

> and supports 511 GB of memory per process

One slimline kernel. 800k.

One vi session. 200k.

One gcc compile. 9000k.

One demand-loaded shared glibc. 3000k.

510.99 GB free virtual memory for a Windows XP install under VMware. Priceless.

Re:Hammer will rock!

[slyfox]

The original Microprocessor Forum presentation is available:
http://www.amd.com/us-en/assets/content_type/Downl oadableAssets/MPF_Hammer_Presentation.PDF

Re:Hammer will rock!

[Sloppy]

Linux has already been ported to the simulator, and supports 511 GB of memory per process. That should do for a start!

Nifty. 511 GB should be enough for anybody.

Remember I said that, and curse my name 20 years from now.

Re:Hammer will rock!

[roguerez]

Too bad Linux actually uses MORE memory than XP with X and Netscape/Mozilla loaded. Or were you gonna run that VMware in text mode?

is there catch phrase going to be...

[geekoid]

...Can't touch this?

Re:is there catch phrase going to be...

[Detritus]

I was thinking more along the lines of "Trust me, I know what I'm doing."

Re:is there catch phrase going to be...

[geekd]

Knowing AMD's past chips, you won't be able to touch it without asbestos gloves.

I guess no one knows yest if this will run at egg-frying temps like past AMD chips.

Re:is there catch phrase going to be...

[geekoid]

I hear that alot, but my 1.4 runs at ~43c. It is rated to ~90c. My ambient case temp is ~32C.
My Pentium 450 runs a ~40c

Re:is there catch phrase going to be...

[fault0]

This is why they invented heatsinks and fans.

Which are recommended even for Intel chips.

Re:is there catch phrase going to be...

[ThatComputerGuy]

Is the bus going to be another of those "double-pumped" deals? If so, the catch phrase could easily be "two pumps and a bump"..

Re:is there catch phrase going to be...

[Anonymous Coward]

I manage to fry a pretty good egg with a Pentium IV as well.

ISA = Instruction Set not the ISA bus

[vondo]

This has nothing to do with what bus is supported. Hammer is continuing and expanding on the x86 instruction set. It has nothing to do with the old ISA (Industry Standard Architecture bus).

Motherboard makers are free (or not) to put an ISA bus on the board. I'd be surprised at the time of Hammer to see such a board, though

The Underdogs

[LighthouseJ]

Why is it everyone wants AMD to pull some magic processor out of it's pants and to kill of Intel? Intel could start a pricewar at any time and price AMD right out of business with a quickness. True, variety and anti-monopolistic practices are good things, but Intel got to where they are because they did some things right. Some people are pulling for AMD because they are the underdogs, others are doing it because they want to jump on the bandwagon.

Those left, those people that say AMD is undeniably better than Intel in all categories are wrong, Intel and AMD have their own set of advantages and disadvantages over the other.

I am reading a tomshardware.com article on how AMD and Intel's previous and latest processors handle heat. The AMD processors failed horribly, they had zero heat protection. Not only are the processors worthless (burned up or not), but the motherboard could be damaged too. They even used a motherboard that the manufacturer guaranteed wouldn't fry an AthlonXP (a brand new processor). Guess what? In less than a second, you wasted hundreds of dollars. I'll just say that Intel's processors, Pentium 3 and 4, they didn't have any damage. You can read about it on tomshardware.com, there is an article called How Modern Processors Cope With Heat Emergencies, they even published their very first lab video demonstrating exactly what they did here.

AMD can't let things like this occur, they have to give customers something that none of the competition can, they need to innovate. Before AMD has enough weight to kick Intel around, it has to have much more support from it's customers. It'll take more than hopes and dreams to push Intel out of the #1 CPU slot. (pun intended)

Re:The Underdogs

[ackthpt]

Yeah, but remember, choice is a good thing. Without AMD nipping at their heels, and spanking them occasionally by getting a faster processor out first (and dodging stupid desperate attemtps like the 1.13Ghs PIII), would Intel be prompted to innovate (granted the P4 just looks like it was the birth of a schizophrenic engineering and bombastic marketing, but I digress, it's something to sell for a while.) In the chess-piece-moving that has become the rolling out of CPU's, the competition has made it interesting and provided some damn fast and inexpensive hardware. Read: Consumers are actually winning, the way it's supposed to be.

Re:The Underdogs

[Cheetahfeathers]

Intel did not get where it is because it 'did things right'. It got where it is because IBM already had a licence with them for other products, so decided it would be cheapest to extend that licence when they looked for what to use for their new PC systems. Intel got a good foot in the door, and has never been put down by a better CPU maker.

Imagine where we would be hardware wise if they had gone with Motorola instead...

Re:The Underdogs

[Anonymous Coward]

Big-endian at 500MHz.

Re:The Underdogs

[geekoid]

Hard to say, the cost of the motorola chips at the time, to the cost of intel(who was just about out of bisiness, and Pratically gave the first chip to IBM), may have stiffeled the begining of the PC.

Re:The Underdogs

[Anonymous Coward]

I don't get the problem. If you pulled the radiator out of your car it wouldn't last very long. I bet the PA-RISC 550Mhz cpu's in our HP-UX boxes would melt as well if you took the heatsink.

Re:The Underdogs

[LighthouseJ]

The difference is if my radiator was made by Intel and it was removed, either the engine would automatically limit the rev's (in the case of P4's) or turn off the car nicely before I threw a rod or burned my rings (Pentium 3). If my car had a radiator built by AMD and it went bad, I wouldn't even have a chance to put on the brakes before my engine block was ruined. If you don't catch onto that metaphor, they should have kept you back in 3rd grade.

Re:The Underdogs

[LighthouseJ]

That's just healthy competition and I'm all for that. It produces better goods and services at a lower price. Don't get me wrong, I'm glad AMD is in the business, I'd support any companies' entry into any industry where they think they have a chance, especially when there's Intel which is much bigger than they are.

As their positions are, Intel could easily make many processors that are bogus, but Gateway and Dell and prefab computer companies will put those in because they are the "new, fastest processor yet!", AMD doesn't have that luxury, every processor they make has to grab at as much marketshare as possible. They are fighting an uphill battle, but I respect them for trying.

You don't need to tell me about the advantages and disadvantages of market competition and capitalism, I already had my Economy classes.

Re:The Underdogs

[LighthouseJ]

Well, hey, life's a bitch, isn't it? But I'm not going to cry over spilled milk.

Re:The Underdogs

[Sloppy]

Why is it everyone wants AMD to pull some magic processor out of it's pants and to kill of Intel?
...
...Some people are pulling for AMD because they are the underdogs

Because success of the underdogs, splits the industry and makes it less committed to any one party. Right now, the 386's instruction set is king of binaries. But in a future world of two mutually-incompatable descendents of the 386 duking it out, software companies will be less able to be able to commit to one or the other.

And if they don't/can't commit to a single instruction set, then they're going to have to deal with their problem some way. Scrapping the idea of native binaries, is one way of dealing with the problem. Ship source that the user has to compile, or ship some kind of intermediate pcode or Java bytecode that is cross-platform. Once the need for binary comformity is broken, then you can buy a real computer and run mainstream software on it.

Or maybe stay with binaries, but accept that you have to deal with more than one. Computer dudes only know three numbers: Zero, one, and many. You can get away with telling your customers "We only support one architecture and if you don't like it, then your money is no good here. We don't want the expense and complexity of dealing with more than one." But once you have to handle the situation of more than one, then you can also handle three or ten. Surely you can see where that could lead...

So back the underdog. AMD's success (and I think they will flourish with this CPU) will hold back progress for a while, but as long as it doesn't completely clobber Intel, and instead they end up splitting the market between them, it could lead, long-term, to progress.

It'll take more than hopes and dreams to push Intel out of the #1 CPU slot.

Yes, and they have it: legacy speed. Gee whiz, you think AMD overheating problem is really a big deal? Consumers have long tolerated silly things like that. If people cared about heat, most people would be running PPC or MIPS right now. They're not. If people cared about short lifetimes of computers, then most would be running something other than MS Windows. They're not.

But you're right, those things matter a little, I guess, so Intel will have some customers. Good. If there's no clear winner, then the winner is us.

Re:The Underdogs

[athlon02]

Actually from what I read from one of my professor's books (Dr. Lipovski is his name and it's his EE345L class at UT Austin), Intel and TI worked on a microcomputer for Datapoint Corp. The system was a horribly slow so TI backed out (keeping some patents on it though), Datapoint got burned & gave up on it too, but Intel was newer then (around 1970 iirc) so they couldn't afford to scrap the idea. The thing eventually sold, with the 4004 processor and today we have P4's. So more than just "did some things right" they got lucky!

AMD began in 1969 (I cooped there one semester last year, so I know it's 1969) and have become leaders in Flash memory technology and also pretty well ranked with embedded processors. So while the K5/K6/Athlon/Hammer lines are well known among many, AMD isn't stuck with just that, nor are they likely to go out of business if Intel starts slashing prices insanely.

Re:The Underdogs

[Anonymous Coward]

Why? Well, I can at least speak about me.

First- Intel is doing NASTY DIRTY THINGS with competition, just like Micro$oft, so at least I support AMD. AMD is no saint too, but playing nicer than Intel.

Second- AMD offers more speed for smaller price, and it runs just fine with linux, althrough I had some problems with windows, motherboard problems i guess.

Third- AMD DOES INOVATE. Athlon architecture is VERY GOOD and INOVATIVE, as far as I understand CPU architectures. I can't speak about Hammer now, but I think it should be a good product.

Fourth- I hate it when a company like Intel rushes a product so crappy as Pentium4 at exorbitant prices, with crappy expensive RAM (made by a company bullying everyone else too), and tries to push competition out of market by threatening Taiwan motherboard manufacturers.

Oh, read more about pentium4 at http://www.emulators.com/pentium4.htm

--Coder

Re:The Underdogs

[maxpublic]

People yammer on and on about the heat tolerance of Intel vs AMD chips as if this were some terribly huge issue, e.g. "what if your CPU fan dies?" In all the time I've worked with personal computers, over at least a couple of thousand machines, never once - not *once*, mind you - have I *ever* seen a CPU fan blow out.

The closest I've ever gotten to a CPU fan failing is when one fell off the processor because the shop that put together the box didn't actually snap it down. Fortunately I heard the 'thump' and turned the machine off. But as for an actual mechanical failure - never.

I don't root for AMD because they're the underdog. I don't root for them at all. I buy from them because right now I can get a 1.4 ghz processor for $149 (retail) that will outperform a 1.7 ghz Pentium. That's the kind of value a smart person just can't pass up.

For that price I'll take the chance on a fan failing. Especially since I've never had one die on me before.

Max

Re:The Underdogs

[NutscrapeSucks]

1) IBM owned 10 or 20% of Intel at the time
2) They started on a 68K machine, but the price point would have been way too high.
3) The PC was originally planned to have an Intel 8080, and used lots of stock parts which were already in common use in the microcomputer business. That equalled time-to-market.
4) Likewise, the 8086 was very compatible with existing applications (CP/M and clones, WordStar, VisiCalc, etc.)

So it was a lot more than luck on Intel's part -- they were positioned perfectly in the market for the PC revolution, and lots of people realized that before IBM did.

Re:The Underdogs

[Anonymous Coward]

In either case the chance of it happening is 0, if you dont get that a device which protects you against something impossible happening is only a competetive advantage because there are too many morons like you and Tom ...

Re:The Underdogs

[dtjohnson]

The posts about the AMD "heat emergencies" are gaining the same status as the posts about "beowolf clusters."

Applications

[1stflight]

Here's the running question as I see it. Will there be enough applications out there compiled to take advantage of the X86-64 arch that will make this a viable processor? As I see it, there isn't allot of call for 64bit Apache, Samba, or KDE. Sorry, I hope they'll pull it off, I just don't see it.

Re:Applications

[rtaylor]

All of teh above run perfectly fine in 64bit compiles. KDE has issues when using the cc included with Solaris, but gcc does the job well enough.

Re:Applications

[fault0]

Were there many 32 bit apps when 32 bit microprocessors were launched? Nope. So we should have never had any 32 bit microprocessors?

Create the processor, and the demand will come.

Yes, someday, you'll see your XFree 5.0 with KDE 4.0 desktop using 2gb of RAM at launch. Suddenly, 32 bit addressing starts to feel really small.

Re:FIRST!!!

[Anonymous Coward]

May all your children drink liquid shit whilst masturbating with steel wool

Itanium, etc.

[ackthpt]

While the thought of Itanium duking it out with Hammer may encourage visions of one company stomping another, plus heated discussions, flame wars, and so on, my interest has always of having a 64bit desktop. Intel some time back indicated that the Itanium was targetted exclusively at the server market, is likely rethinking that point. Perhaps McKinley (the joint project with HP) is Intel's idea of the post P4 desktop processor, as I've seen elsewhere that Itanium's x86 emulation makes a PIII look attractive.

The ability to build a desktop workstation with the ability to run all my old x86 crap, fast, and move into 64bit software, also fast, is highly attractive. Athlon or P4 will undoubtably be the choices for the next year, but when AMD gets the Hammer out into the mainstream with a mainstream price, Intel watch out.

Lastly, Microsoft, last I read, didn't indicate any interest in doing a version of XP for the Hammer. Perhaps that hasn't changed. If not, there's a potential hole through which someone may exploit Microsoft's disinterest. Linux, sure, AOL, Hmmm, you know that's a mean fight going on between Reston, VA and Redmond, WA, if the Hammer is attractive to home users, don't be surprise if AOL chooses to support it. It's entertaining to think about, anyway, however you feel about the combatants.

Re:Itanium, etc.

[cymen]

What exactly would AOL do to support a 64 bit processor?

Re:Itanium, etc.

[Glock27]

While the thought of Itanium duking it out with Hammer may encourage visions of one company stomping another, plus heated discussions, flame wars, and so on, my interest has always of having a 64bit desktop. Intel some time back indicated that the Itanium was targetted exclusively at the server market, is likely rethinking that point.

Itanium isn't just for the server market now. IBM, SGI and several others are marketing Itanium technical workstations. Intel has also stated that it sees Itanium making it to the desktop at some point in the future, replacing x86.

Hammer, on the other hand (specifically Clawhammer) has always been targeted at the desktop from the get-go (along with server and workstation). Check it out on the AMD processor roadmap (which I just managed to find again;).

Another point to keep in mind is that the ability to compete in the server marketplace is a key for AMD. It will provide them with the same ability as Intel to subsidize desktop processors with expensive server processors. Right now Intel can sell P4s at a loss and still turn an overall profit, while AMD suffers. Once Hammer ships, the dynamic will change quite a bit... ;-)

Perhaps McKinley (the joint project with HP) is Intel's idea of the post P4 desktop processor, as I've seen elsewhere that Itanium's x86 emulation makes a PIII look attractive.

I thought McKinley was just the .13 micron version of Itanium, perhaps with more cache. Does it have an enhanced ability to do IA32?

The ability to build a desktop workstation with the ability to run all my old x86 crap, fast, and move into 64bit software, also fast, is highly attractive. Athlon or P4 will undoubtably be the choices for the next year, but when AMD gets the Hammer out into the mainstream with a mainstream price, Intel watch out.

I couldn't agree more!

Lastly, Microsoft, last I read, didn't indicate any interest in doing a version of XP for the Hammer. Perhaps that hasn't changed. If not, there's a potential hole through which someone may exploit Microsoft's disinterest. Linux, sure, AOL, Hmmm, you know that's a mean fight going on between Reston, VA and Redmond, WA, if the Hammer is attractive to home users, don't be surprise if AOL chooses to support it. It's entertaining to think about, anyway, however you feel about the combatants.

I think Linux will be strong presence on the Hammer, along with potentially (wild prediction here) MacOS X. Microsoft will support it as soon as it begins to take marketshare like the US Rangers taking Omar's palace (not that I particularly care if Microsoft supports it). As for AOL, it should just get busy porting it's interface to Java like it said it would a year or so ago. That alone would be a big blow to Microsoft, and would simplify software development quite a bit for AOL as well as widening the number of AOL platforms substantially.

299,792,458 m/s...not just a good idea, its the law!

Re:Itanium, etc.

[csbruce]

AOL, Hmmm

It seems to me that power users and businesses would have most of the interest in using 64-bit processors.

AOL's target market probably has more modest requirements and maybe AOL should be looking into buying up XBoxes, loading them up with Linux and Mozilla, and selling them as set-top surfer boxes.

Re:Itanium, etc.

[csbruce]

maybe AOL should be looking into buying up XBoxes, loading them up with Linux and Mozilla, and selling them as set-top surfer boxes.

Actually, they could just distribute millions of CDs that do that.

Re:Itanium, etc.

[Anonymous Coward]

Itanium = Designed by Intel, not too good.
McKinley = Designed by HP, is to be what Itanium was meant to be.

Re:Itanium, etc.

[maraist]

I thought McKinley was just the .13 micron version of Itanium, perhaps with more cache. Does it have an enhanced ability to do IA32?

McKinley is a whole mess of add-ons.. Not least of which is the idea that it can issue more EPIC instruction / clock than the Itanium. The original idea was that Itanium would chapion the instruction set, but would be an unwieldy beast with all it's new features.. But it would be enough to transition the market place (too bad it's practical performance sucked). McKinley would then be the knock-out punch that fully utilized it's potential (though at greater cost due to increased numbers of components). From this Itanium would be a low end that allowed "entry-level servers". Then they'd have time to go redesign new features for their next [incremental] generation... Their EPIC instruction set has templates so that adding whole new classes of functionality "should" be trivial.

Course I don't think they expected having to relegate Itanium as a "pilot" CPU with embarrasingly low frequency ratings (but MHZ is all that matters, right Intel?). Doesn't sound like the P4 guys are under the same marketing department as the Itanium guys (GM in the making?)

-Michael

Re:Itanium, etc.

[fault0]

Actually, an Itanium's x86 compatabity mode makes a PII or Pentium Pro look attractive.

Re:Itanium, etc.

[Anonymous Coward]

AOL actually did try a Linux/Mozilla settop box. Sold like crap.

Re:Itanium, etc.

[Jordy]

my interest has always of having a 64bit desktop.

And you need access to 16 exabytes (or 8 w/ signed pointers) of address space in your desktop applications because....? (not total memory, but memory per application as you can have more than 4 gigs of memory on a x86 processor in a single machine.)

I don't know where this idea that 64 bit memory addressing makes programs run faster came from, but there is nothing inherent about 64 bit addressing that would make it faster for your average integer based desktop applications.

Of course, I guess it all depends on your definition of a "64 bit" chip architecture. I tend to define it as an architecture's registers, data bus and ALU are all 64 bits wide.

I don't know about you, but unless I need more than 4 gigabytes of memory per process or I'm doing some heavy floating point where I need 64 bits of precision, I don't particularly want my data structure heavy applications using up to twice the memory they used to.

Of course that's just my opinion; I could be wrong.

Re:Itanium, etc.

[ackthpt]

It's the registers. 64 bits are more sexy than 32 ;-)

Seriously, it's the work I'll put the CPU to, not the memory that makes these attractive.

Re:Itanium, etc.

[be-fan]

One nifty thing about 64bit memory addressing that is often missed is that it makes OS design tons easier. First, you can just contiguously map all of physical RAM permanantly instead of dynamically mapping in needed regions (like Linux high-mem). Even on many desktop machines, people are coming up to Linux's 1GB kernel-space address limit and having to use the more complex highmem code. Also, the 4GB address space of 32bit procs can become exteremly limiting when you have to deal with memory mapping large files and such. Lastly, library management becomes tons easier. Usually, libraries on 32 bit systems have to be relocated because the bases of their compiled images can conflict with those of another library. On a 64-bit arch, it is feasible to assign each library a unique base address and never have to relocate after the first time.

Re:Itanium, etc.

[Chep]

>>Lastly, Microsoft, last I read, didn't indicate any interest in doing a version of XP for the Hammer. <<

While they have officially announced nothing (after all, they're sort of a partner of Intel especially regarding the Itanium, aren't they ?), and AMD officials insisted on saying nothing on behalf of MS (but according to The Vulture's journalists, that wasn't without a little smile),
they did sprinkle the DDK sources with Hammer-related conditional symbols.

Re:Itanium, etc.

[Lars T.]

5 more, and things get interesting ;-)

Not really

[vondo]

So intel's plan to bolster their flagging market share is to introduce an entirely new platform that's not backwards compatible? I'm not exactly hardware-techno-geeky, but it seems to me this means none of your software is going to run (It would be like running a mac program on windows.) It seems to me there's got to be a way to have an intermediate step somewhere down the line that support both architectures. But then again, I could be an idiot on this point.

Itanium/McKinley will be backwards compatible with the Pentium series, just not very fast since it emulates the instructions.

Remember, Apple did something similar with PowerPC except it was emulated in software. What happens is, the OS and the major applications (Office, Photoshop, etc.) are ported over to the new instruction set. Then, who cares if your Winamp is actually running in emulation mode? Most of what you do is running in native mode.

The Itanium is targeted at servers first where a limited number of apps are needed. Then expect to see it on business desktops with a few more apps and finally percolating down to the home user in 2004/2005 (?)

Re:Not really

[ivan256]

As far as I knew McKinley was not going to have backwards support. IA-32 support is an optional part of the IA-64 spec, and probably only Itanium will implement it.

Also, Apple's software emulator ran the 68k code on PPC at speeds that were roughly equivalent to the 68k. Itanium, when running 32bit and 64 bit programs at the same time, performs very poorly. Itanium also does not have the benefit of the MacOS engineering team that did a remarkable job making the transition seamless...

I hope that Intel finds a way to reduce the power consumption of their 64bit chips.

Re:Not really

[Anonymous Coward]

The real answer is that they all have IA32 compat. The Itanium does stuff like out-of-order scheduling to the IA32 micro-ops, while the McKinley does not. It was done this way b/c the Itanium was damned good at running IA32 instructions, but it was so complex that it slowed the whole processor down. The McKinley does away with that and puts a straight-issue IA32 emulation on board which allows for similar IA32 performance levels and kickass IA64 performance levels. That way, we have our cake and eat it too (IA32 compatiable new architecture!)

Incredible submission

[Mahtar]

Have anandtech more information on grammar/sentence fragements?

Re:Incredible submission

[Anonymous Coward]

Appears have not information for such said inquiry.

Re:Incredible submission

[Anonymous Coward]

extremely the funniest reply have i read in days!!

This post is clearly offtopic too

[Anonymous Coward]

Mod it down.

Re:This post is clearly offtopic too

[Anonymous Coward]

Wow, only one-liner, auto -1 mod!

It's all in the UID...

[Anonymous Coward]

With a UID of 47, this must be a post from an insightful member. After all, he was here the day that UID's were first issued!

The need will come.

[Kjella]

Where's the need for 64-bit desktop chips?

Somehow I got a feeling this was asked before.. like "Where's the need for 32-bit desktop chips?" or "Where's the need for 16-bit desktop chips?".

Even if we can't find any killer apps that _need_ 64 bits today, we'll figure something out. Maybe 4gb of ram limit, I "only" got 3000 times more ram now than when I started with my 64kb C64, and things seems to eat up more and more, not to mention processing power (and don't get into bloat, even linux needs more and more of both), which can be achieved two ways:

Faster processors (clock cycles, parallellization), or more work pr. clock cycle (64bits vs 32bit, new instructions).

Kjella

Re:The need will come.

[Anonymous Coward]

The x86 has a 36-bit addressing mode so the RAM limit is 64 GB. It's the operating system support that is sticking people to 4 GB. Higher end MS stuff can currently do > 4 GB and some Linux and BSD variants can as well.

Re:The need will come.

[NutscrapeSucks]

Where's the need for 32-bit desktop chips?

That's exactly the problem I'm talking about:

Intel: Here's a cool 32-bit chip, somebody write some software.

Microsoft and IBM: We don't need to support 32-bit for the next 10 years, so you get a bunch of crappy compatibility hacks and spurious "out of memory" errors. The hacks will make it _more_ difficult to support 32-bits in the future. Enjoy!

Now, the exact same thing is going to happen all over again for 64-bit chips. And I'm supposed to be excited?

TROLL!

[Anonymous Coward]

It's just another adequacy.org troll...

Troll recognition made easy:

[Anonymous Coward]

1) links to advocacy.org
2) mentions socialism
3) you need another? see (1) advocacy.org is a pure trolling machine

Alpha? You probably don't even run VMS...

[Anonymous Coward]

You heathen savage!

Re:Alpha? You probably don't even run VMS...

[Anonymous Coward]

Ha! You VERIFYable whore of Satan. I have VMS but only hobbyist, as our business hasn't quite standardised on VMS yet :-).

Anyway, what happened to programming in assembler? Ah yes, I remember, it got so ugly with IA-32 that it became a forgotten skill.

FUDpacker...

[SaDan]

Buy a heatsink, you cheap bastard, and install it.

name?

[TI-83]

The article says something about getting "... very high clock speeds (10GHz+ by 2006)..." Wasn't there a /. article last week (or so) about new transistor packaging (within the chip) that would be boosting intel speeds to 20GHz by around 2006? I'm sortof wondering which will emerge as an actual product. and how the heat issues and stability will be managed.

AMD Technical Information

[Coniine]

Maybe someone here can help : I may be doing something stupid but I can't get any technical datasheets from the AMD site for things like FLASH, CPUs, Chipsets. It appears that I need to contact AMD and go for some sort of NDA...

Just another extension?

[Innominandum]

It looks like AMD are extending the same architecture (IA16->IA32->Hammer) again. For developers it's tempting to stick with what's familiar. But there are still "only" 16 float and 16 general-purpose registers. There isn't a big change when compared to IA64. This looks like another processor based on the over-developed, simple, tried-and-true designs that hail from the 1970's.

I am not a big fan of Intel, but at least they are addressing the deficiencies of IA32 architecture. They are replacing it and not adding to it. The IA64 design isn't amazing by any means. But it's our best hope for a "next generation" processor.

Why don't AMD clone Itanium? Considering their current position, I think it would be the smartest move for now.

Re:Just another extension?

[cmowire]

It's likely that the K9 (I hope they have a dog or dental-related sort of codename for it. ;) ) will be dependent on how the Itanium and Hammer do on the market.

If the Hammer cleans up, the K9 will build on it, leaving any possibility for a whole new platform to the K10. If the Itanic architecture starts to gain speed, the K9 will probably be an IA64 machine.

I think the key thing is that the instruction sets are mattering less. You can put optimizers in hardware that convert a messy x86 architecture to a nice RISC one. Think of the x86 arhictecture as a compression format for nice RISC opcodes. Or you can do various kinds of software morphing, which are getting more advanced as time goes on. The only real advantage to the IA64 is that it has the likelyhood of allowing the compiler to make better optimizations that will leaverage the processor more.

Re:Just another extension?

[Glock27]

I am not a big fan of Intel, but at least they are addressing the deficiencies of IA32 architecture. They are replacing it and not adding to it.

Are they addressing the deficiencies? By what metric do you make that claim?

The IA64 design isn't amazing by any means. But it's our best hope for a "next generation" processor.

How so? So far Itanium performance has been less than impressive (to put it kindly).

Why don't AMD clone Itanium? Considering their current position, I think it would be the smartest move for now.

AMD is trying to differentiate itself from Intel by providing superior price/performance, raw performance, and backward compatibility. So far Itanium has been a big disappointment (don't forget it was originally supposed to ship in 1998, IIRC). One important factor will be how Hammer die size stacks up against Itanium die size using the same process. Anyone have a clue?

299,792,458 m/s...not just a good idea, its the law!

Arrggg. stupid moderators!

[recursiv]

Not funny! Not funny!

x86 sucks.

[Anonymous American]

x86 isn't dead yet, and IA-64 will never live!

Umm, and this is a good thing? The x86 instruction set sucks. I can't wait for EPIC (or something better).

Link correction

[Anonymous Coward]

Sorry, I must be drain bamaged :) Here's the fixed link, in case you wanted to check it out. It's like slashdot, but oriented to hardware more.

Double-check your assumptions

[acidblood]

Itanium will have a run for its money, I suspect.

For that to happen, first those two processors must be in the same market segment. Do you see people using Alphas switching to x86? The same way, Itanium's market is not the desktop or average workstation. They're trying to compete with the cream of the crop RISC chips.

Rants aside, the Hammer architecture is a nice way to transition to 64-bit computing while keeping legacy code, no rewriting or even recompiling. But it has a cost: it'll carry the same inefficiencies found in current CISC processors. While many designers have found their way around it in the performance arena, the drawbacks are evident: higher complexity, power dissipation, etc. -- die size is unevenly divided between control logic and number-crunching logic, which isn't a comfortable situation.

Take the time to generate assembly code using gcc -S, and you'll soon find out there's just a handful of instructions being generated. Taking a peek at AMD's code optimization guide, you can see those are the fastest instructions (collectively called DirectPath instructions), and they can be paired with 2 other simple instructions in the decoding stage. Trying to execute hand-optimized 8086-80386 code would actually result in slow execution, due to the use of VectorPath instructions, which can't be paired and, as a rule of thumb, take more cycles to execute (not always the case though. We're RISC-ing the code in order to make it efficient, but we're carrying the burden of Intel's design flaws of long ago, in the form of this added control logic for handling legacy code.

At this point, since rarely used instructions are being executed slowly, why not resort to some sort of emulation, a la Code Morphing? Not completely throwing away the architecture as Transmeta did, keep the simple instructions and emulate the rest (such as those used to handle BCD arithmetic, hardly used today).

Although it may seem like I hate x86, in fact, as an assembly coder, I find it one of the most comfortable architectures to work with. I have read some RISC assembly code, and couldn't picture myself writing code in it. I've programmed in 68HC11 assembly as well (it's a microcontroller from Motorola), and it feels like coding with your hands tied. So I'll definitely enjoy the new Hammer archicteture, especially because it fixes my two major complaints: small amount of registers (they've added 8 new ones), and the 32-bit architecture. Also, I'll be drowning for SSE2 support, including 8 new registers as well (that's 16 128-bit registers! yay!)

Re:Double-check your assumptions

[Awxxx]

My grand grand father's first car had only one gear : forward. Right now, he owns a car with 6 speeds gear box but he is stucking using the first one : why change the "good old driving method" ? You can tell me that automatic gear shifting does exist but why are those car racers using manual gear box ? to gain weight on the car ? Right now we are using automatic gearshifting, can't we go farther ? If computing is not about going faster and faster, why bother building new CPUs ?

Re:Double-check your assumptions

[roca]

> keep the simple instructions and emulate the
> rest (such as those used to handle BCD
> arithmetic, hardly used today)

In fact, in Hammer's 64-bit mode, the BCD instructions (and some others) are not supported.

> the drawbacks are evident: higher complexity,
> power dissipation, etc.

Check out the heat dissipation on Itanium! One guy I know has a box that puts out 120W per CPU.

A simpler architecture is a nice thing, but experience seems to have shown that it doesn't matter that much in practice.

Re:I'm very interested

[Anonymous Coward]

Get some more information here. I'm not even going to bother hiding that one.

But can it match the G5's SpecMarks?

[GPS Pilot]

SpecInt2000: 987 @ 1.2 GHz, 1340 @ 1.6 GHz
SpecFP2000: 1005 @ 1.2 GHz, 1359 @ 1.6 GHz

(source: The Register)

I don't expect AMD or Intel will come out with anything comparable until 2004.

Re:But can it match the G5's SpecMarks?

[Anonymous Coward]

> I don't expect AMD or Intel will come out with anything comparable until 2004.

Which is still 2 years before the G5 will hit the markets..

Re:But can it match the G5's SpecMarks?

[Anonymous Coward]

Expect the processor to be released 01/2002. Expect computers soon after, and expect a huge demand for them.

Re:But can it match the G5's SpecMarks?

[Anonymous Coward]

The only thing I expect is to be disappointed, AGAIN, as Apple squanders yet another technology advance.

Re:But can it match the G5's SpecMarks?

[Anonymous Coward]

Hammer SpecInt2000 1400 @ 2.0 GHz
according to AMD

greetings fellow vax 2000 user!

[Anonymous Coward]

How is the 500mb disk drum array working?

Upping the memory on our VAX to 256mb allows us to handle 20 concurrent vt100 users.

Have you any information on the vt680 terminals?
Rumor has it that they run at 14k baud!

I just got and love the ANSI C complier on this sytem. It handles block i/o with the greatest of ease.

Take care.

Where there's goatse...

[Anonymous Coward]

Hey, the mods got one right. Not only does your post deserve (-1, Redundant), it also needs (-1, Retarded), (-1 Cum-Guzzling), and (-1, Nigger Post). Good job!

Speaking of Hammer...

[Anonymous Coward]

Who controls the British Crown?
Who keeps the metric system down?

We Do!
We Do!

Who leaves Atlantis off the maps?
Who keeps the Martians under wraps?

We Do!
We Do!

Who holds back the Electric Car?
Who makes Steve Guttenberg a Star?

We Do!
We Do!

Who robs cavefish of their sights?
Who rigs every Oscar night?

We Do!
WE DO!

What about the pricing?

[c.r.o.c.o]

There isn't any mention in the article about the expeted prices of the Hammer, so I thought I'd ask here. What are the price expectations for a processor like this? I mean from the specs alone (with so much stuff integrated into the die), it's going to be a fairly big beast.

Does the fact that it is new technology, and that it's a big (or bigger) die size automatically mean it's going to be very high priced? I remember when the P2 came out, I paid CAD $1200 for the 300Mhz, about 2 weeks after it was released. Now the P4 costs about the same (although a bit less than that) for the highest speed (2Ghz?)

So my question is this: will this processor be affordable (somewhere between a top of the line Athlon and a P4), or is it going to be much more? I think it's a very safe bet to assume that it will cost more than the Athlon.

If somebody has a real answer for this, please reply. It would be interesting to hear some opinions from the more knowledgeable.

Re:What about the pricing?

[s390]

What are the price expectations for a processor like this? I mean from the specs alone (with so much stuff integrated into the die), it's going to be a fairly big beast.

Expect AMD's Hammer chips to cost much less than Intel's Itanium CPUs. Intel spent several years and likely Billions to develop the Itanium, whereas AMD needed only about one year for the Hammer. The first model will be the Sledgehammer, targeted to Servers, so those won't be exactly cheap. But the second model Clawhammer CPU will be for Workstations/Desktops and probably comparable in price to current high-end Athlons.

Re:What about the pricing?

[RedWizzard]

The price will be dictated by the market. The first Hammer release will likely be the server version SledgeHammer which will be priced to be competitive with Itanium and P4 Xeons. The desktop version ClawHammer will start out pricey as AMD look to clear existing Athlon stock. It'll probably still be at a similar or slightly lower price than the top of the range P4s at the time. It is meant to be a replacement for the Athlon so it will eventually need to be priced accordingly (i.e. cheap) to suceed.

history of tech

[dollargonzo]

aaah, the golden days of technology have come and passed. Before, the technology ppl used, they actually could see and understand / know the functioning of the device. Because of this, repair was almost trivial, and technology was much less of an elite ordeal than it is now.

Over the years, tho, technology has progressed to the point where the ppl DESIGNING the damn stuff, for god sakes, dont even know half of how th product or tech functionsm, besides being able to fix it. If we were to have a technological apocalypse, it would take an enourmous amount of time until we were at the stage that progress could be made again, EVEN if we still had the resources / knowledge of the products we are making.

On another subject, ppl using tech are so blinded by what they call AMAZING speeds, performance, etc, and ask how AMD can possibly surpass what they have currently made....the question is simple: make newer better, and faster stuff!!! we should not be marvelling @ all the NEW things that
ppl come up with, but rather the rate of advancement. back in the day, the 1khz machine was considered a lightning speed, and held in an enourmous warehouse. Ppl should understand that innovation is ALWAYS going to happen, and not wonder at how something could possibly be better / blindedly speculate about the future.

remember, there was a time when ppl didnt believe that widespread electricity was ever going to be a reality, but we ALL know, that the next technology is ALWAYS 5 years down the road...

Nope, you're retarded

[digitalunity]

It's all there. You just have to dig. You can get a white paper for just about anything on their site.

How stupid are you?

[Anonymous Coward]

the idea is to shut down the /. community to pressure CmdrDickweed to make the system less opressive

I guess you're not bright enough to realize that if you weren't so stupid then the "opressive" measures wouldn't ever have been put in place to begin with?

Maybe by attempting to "break" /., you'll only get more "opression"?

Re:How stupid are you?

[Anonymous Coward]

Oh yeah? Show me the posts where I caused the ip ban idea to be implemented, jackass. Show me where I caused content filters to be implemented, or wait times, dipshit. We're both posting as AC, so if you want to blame someone, you're just as culpable as I am, dumbass fuckwit. First of all, I'm not involved with this, I'm just trying to guess the motivation. Second, you must have the IQ of a turd. If everyone who views Slashdot gets IP banned, then they'll obviously impliment a better banning system, one that doesn't take out groups of people for the actions of one. Retard.

Re:How stupid are you?

[Anonymous Coward]

Oh, yeah?

Nice comeback, Potsey!

Poor little dipshit, got his panties all in a knot.

Awwww......

Re:How stupid are you?

[Anonymous Coward]

If it weren't for posters like you with a vocabulary as low as your IQ, there'd be no need to have an IP ban - there'd be no intentionally bad posts. There'd be uninformed ones that would promptly be modded off the map, but no trolls, no flames. In an ideal world, of course.

Even in a less ideal world, there'd be few trolls, and it wouldn't be worth the trouble to do anything more than mod them down.

Because of the higher level of trolls, *something* had to be done. It is obviously a stop-gap measure to keep out the worst of them. Since you're posting anonymously, I can't show you the posts where you caused the ip band idea to be implementd, jackass, but if you've trolled or flamed before, you probably had a part of it.
There's already a system out there. It's imperfect, but they all are. Why make things harder on the slashcoders than they have to be? Why not submit a patch yourself if you're so concerned, rather than sit around and bitch about it and try to figure out ways to abuse it. That's what's causing these measures to come about in the first place, and while it's obviously not working, what the hell do you want the slashcoders to do? They've obviously had complaints about the trolls, and they're desparately seeking a way to prevent that.

If it becomes a problem, I'm sure it will be removed from the code, but it seems to me that the only problem likely to come of this is from trolls abusing the system. Why bother?

Ummm...too stupid to think of a way to close my comment...Retard.

Re:How stupid are you?

[Anonymous Coward]

blah.blah blah-blah blah

blah blah/blah blah=blah blah:blah blah blah

blah;blah'blah blahblah blah,blah

blah1blah blahblah blah3blablah blahblah blah5blah

blahblah blahhblah blah8blah

retard....

Re:How stupid are you?

[Anonymous Coward]

Touchè

Re:How stupid are you?

[Anonymous Coward]

Touchè my monkey!

govt == stagnation in technology

[Anonymous Coward]

retire please, oh 1970s unix academic lamer

Maxwell?

[xanadu-xtroot.com]

Joan was quizzical
Studied patophysical science in the home
Late nights all alone with a test tube, Oh-oh-oh-oh
Maxwell Edison
Majoring in medicine calls her on the phone
Can I take you out to the pictures, Jo-o-o-oan?
But as she's getting ready to go a knock comes on the door


Bang Bang Maxwell's Silver Hammer came down upon her head
Clang Clang Maxwell's Silver Hammer made sure that she was dead


Back in school again
Maxwell plays the fool again, teacher gets annoyed
Wishing to avoid an unpleasant sce-e-e-ene She tells Max to stay
When the class has gone away, so he waits behind Writing fifty times I must not be so-o-o-o
But when she turns her back on the boy
He creeps up from behind

Bang Bang Maxwell's Silver Hammer came down upon her head
Clang Clang Maxwell's Silver Hammer made sure that she was dead

Re:Maxwell?

[talks_to_birds]

oh..

*I* get it...

"hammer"

huh.

t_t_b

Hell no!

[Nindalf]

Don't you think that we've hauled along the old 8086/XT baggage long enough?

Hauling baggage builds character, consarnit! I was born with a new x86 chip on the market, and I'll die with a new x86 chip on the market.

Some day, I aim to teach my great-grandchild to xor the accumulator by itself to save a byte off the executable, so don't you go a-messin' with that dream!

MOD UP!

[Anonymous Coward]

Pretty funny. Probably unintended by the poster who just meant to troll :-)

Re:MOD UP!

[Anonymous Coward]

Yeah. I suppose actively porting to new architectures constitutes a dying platform.

x86 code on Itanium

[Dr. Spork]

Yeah, I've seen the benchmarks. An 800 MHz Itanium is trounced by a 133 MHz Pentium when it comes to running x86 code. This hardly passes for backward compatibility.

Re:x86 code on Itanium

[Anonymous Coward]

and to think of how much effort and silicon went to doing this stupid harware emulation. amazing.

Re:x86 code on Itanium

[default luser]

This is especially bad when you look at how efficiently Transmeta's Crusoe manages to emulate x86 code...wow, an IA-64 core with code-morphing, what a concept!

The Hammer Creed:

[Morbid Curiosity]

Marketing is our shield
that protects us from the competition.
Benchmarks are our weapon,
with which we carve a path to an overclocked future...

..OK, perhaps I've played Thief too much. I guess I'm too illegit to quit.

I will use it.

[SaDan]

As soon as I can buy processors and motherboards to build a cluster, or a few standalone high end multiprocessor machines.

AMD's got a great product in the Athlon, and I can't wait to use the Hammer.

Fine performance is fine...

[Guillaume Ross]

But will it burn like an Athlon?

32/64 mixed environments

[captaineo]

I'm really curious to see how things are going to work out on x86-64 and IA-64 with regards to mixed 32/64-bit systems. I assume both platforms are going to provide some sort of thunking interface so 32-bit code can call 64-bit libraries, etc, but it's a non-trivial problem, especially for developers, since you have to devise some system for building and keeping both versions of libraries and executables around.

Sun and IRIX people have been dealing with this for a long time, but I'm interested to see how most other developers will adapt (since I'm guessing there are 20-30x more x86-only developers than Sun and IRIX developers =).

Crap flood v8.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v8.0

Because Itanium is not proven

[Anonymous Coward]

That's right, it is still regarded as expermintal architecture. People are still trying to figure out the compiler optimizations that will allow massive parallelization. Basically it's a superscaler vs vliw battle (yes, epic is basically vliw) at the moment superscaler processors can still have the performance advantage, maybe that will change in a few years time but it wont be any time soon.

Crap flood v9.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v9.0

Crap flood v10.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v10.0

Crap flood v11.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v11.0

Crap flood v12.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v12.0

Crap flood v13.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v13.0

AMD doing good, but

[A_Non_Moose]

someone has to say the "M" word...marketing.

I have yet to see an AMD commercial, and word of mouth (yes, even mine) only carries so far.

AMD processors are simply increadible, IMO, but how to get the word out? Marketing, commercials and ads.

It is a simple question, really. What is the point of having such a great processor, if no one knows it?

I think a simple commercial like this would work wonders:
Open on a little tv playing the p4 "blue man group" commercial....have a "sledge hammer" and a "claw hammer" (both with big AMD stickers) smash the tv into oblivion.
(fade to black with the AMD logo and a "well known voice")
The AMD Hammer series and XP series, smashing 'you know who's higher numbers".
Power is *sexy*, AMD.

Or, as a demo, us the the ending of "The fast and the Furious' " car race.
Amd would be the Black Toranado(?) and Intel the Honda(?)...Raw Horsepower vs high rpm and technology+"cheats" (inflated Ghz = NOS, perhaps.)

Essentially, it was a tie.

Draw your own conclusions, or come up with something better.

Moose, out.

Re:AMD doing good, but

[Anonymous Coward]

AMD did run one commercial when the Athlon was released. Very cheap looking - the plot was if you run a Pentium, a train will smash through your wall and kill you. I doubt it was very effective.

Of course one big reason that Intel is more expensive is that they give big co-advertising kickbacks to the OEMs (which is why you see Intel Inside all the time on OEM ads).

Also, a company's fanbase always wants commercials which says "Ours Is Better Than Yours, Nah-nah!", but that's preaching to the converted and not always good marketing.

Re:AMD doing good, but

[ThatComputerGuy]

No, NOS is actually very potent power-adder, as it basically does the job of any forced induction device such as a turbo or blower.

I think a comparison of inflated ghz = extremely high redline w/o adequate power through the whole range.

Take the Honda S2000, for example: small, 2.0L engine, amazingly powerful though. It makes 240hp out of those 2 tiny liters. What does it rev to? 9000RPM. Sounds great, no? NO! I sure hope you enjoy driving at 6000RPM constantly to get any passing power, because that's where the power band lies-- some range between 6000 and 9000.

So what's the use of the rest of the rev range? You tell me.

Crap flood v14.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v14.0

Crap flood v15.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v15.0

Point of view from a electronic/computing engineer

[Awxxx]

I had to do a review of IA64 and I wanted to know what was AMD's response to Intel 64bit CPU and what was behind the "old" generations.

Currently, 2/3 of a CPU is used to analyse/understand/reschedule the code send to the CPU. This part is very important and AMD seams to be better at this game than Intel. The code has to be reschedule so the different parts of the CPU that can work at the same time are efficiently loaded ....

OK, let stop right now : why isn't the code already efficient ? Because the compiler does NOT care about the inner structure of the CPU so the CPU has to do all the real work.
By keeping with the "good old architecure", AMD is trying to do in hard and in real time what a software (let's say a compiler:)) can do much more easily in a very long time. And a CPU can't see more than a few operations ahead whereas the compiler can see the WHOLE code.

So, by removing all the optimisation crap from the CPU and showing the compiler what's reallly inside, Intel is on the right way. In current CPU, you have more than 40 registers, but you can access only 8 of them and the CPU has to "guess" what could be the best use of them.

So, I think Intel's approch is the right one. Just recompile all your software : to run old stuff, use old hardware.

I have datasheets and documents to comment about this and I would glady do it.

Crap flood v16.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v16.0

Speak of the Dinosaur

[Anonymous Coward]

Man this guy's got such a l33t UID, I'm afraid to even mod him.

Exactly.

[digitalunity]

Top three reasons to use AMD:
1: Support the underdog
2: Try something new and inexpensive
3: Test the fate of the gods

When you look at AMD's history, they've always been the follower; attempting to deliver what Intel has always charged too much for. The Hammer architecture is a serious departure from the 'follow the leader' approach they've used in the past. Even the Athlon was not nearly so different from the Pentium as the Hammer is.

In fact, the Athlon was nothing more than a Pentium Pro on steroids. The Hammer is a radically different beast. The smooth transition it will allow for developers will be something the average user can appreciate.

IA-64 will take a long time to become a standard, if ever. I think it will find it's home in big Blue boxes, and Sun enterprise solutions. The only immediate uses are going to be scientific and industrial. Both good markets, but Sun, MIPS, HP, Compaq(Alpha as well) have proven through history that big business alone cannot support an architecture. You need the large installed user base that the Pentium got.

When examining a processors performance, you have to look beyond it's theoretical performance, and understand why they fall so short of the expectations. Even under ideal circumstances, the Athlon just cant keep busy. The ALU's just don't stay full. With an integrated North Bridge on every Hammer. you end up with an far decreased latency and better scaling with processor speed. Closer inspection also reveals a mechanism to link every Hammer to up to two other Hammers. The memory attached to each Hammer is unioned so that the OS sees one big memory map.

In the development of microprocessors, most ideas are nothing more than the next logical step; an slightly innovated version of the last. The Hammer tempts us with a revolution. Processors that stay busy all the time. Processors that talk to each other directly. Some say that a built-in memory controller is a Bad Idea. Being built-in leads to higher performance. Much lower latency as well.

In the past, revolution was only possible by visionary minds. AMD has those minds. AMD's ability to execute has risen my many orders of magnitude in the last 3 years. If they can keep it up, Intel may lose their crown(and their checkbook). Personally, I think they can pull it off. I also believe the Hammer will be received with much greater market acceptance than both the IA-64 or the P4. If for nothing else but competition's sake, I hope the Hammer succeeds.

Crap flood v18.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v18.0

Crap flood v19.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v19.0

Crap flood v20.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v20.0

Re:Crap flood v20.0

[Anonymous Coward]

ip taliBAN this slut whore!

What goes down, must come up

[The Living Fractal]

Watch the AMD hype, it's going to get pretty heavy in the near future.

People are already choosing AMD as their champion of Intel, and they're also already pulling their selective strings to hype up AMD processors that aren't released or benchmarked.

Can you said biased?

/rant mode on

These people are just like the company Apple. They have a common slogan, 'think different'.

If everyone thought 'different' and bout Apple, everyone would be thinking THE SAME WAY.

stupid &$@^s

/rant mode off

Crap flood v21.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v21.0

Crap flood v22.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v22.0

Crap flood v23.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v23.0

Crap flood v24.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v24.0

Crap flood v25.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v25.0

Slashdot requires you to wait 2 minutes between each successful posting of a comment to allow everyone a fair chance at posting a comment.

It's been 1 minute since you last successfully posted a comment

If you this error seems to be incorrect, please provide the following in your report to SourceForge:

Browser type
User ID/Nickname or AC
What steps caused this error
Whether or not you know your ISP to be using a proxy or some sort of service that gives you an IP that others are using simultaneously.
How many posts to this form you successfully submitted during the day
* Please choose 'formkeys' for the category!
Thank you.

VLIW will fall by the wayside

[Anonymous Coward]

SMT is the way forward to get best performance out of single processor cores, for multiple cores CMP is an orthogonal concept which will also have to be used to make use of the ever increasing number of transistors available, and irony of ironies good old x86 is actually a better vehicle for SMT precisely because of its small register set.

I wouldnt say its an optimal instruction set for SMT processors, but certainly better than VLIW (which makes most sense if the ISA is tightly bound to the underlying implementation).

Diminishing returns for getting more performance out of cores executing a single thread began hitting hard a long time ago, very wide VLIW was simply a step too far IMO. VLIW makes sense in the embedded realm, not desktop processors.

Crap flood v26.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v26.0

Slashdot requires you to wait 2 minutes between each successful posting of a comment to allow everyone a fair chance at posting a comment.

It's been 1 minute since you last successfully posted a comment

If you this error seems to be incorrect, please provide the following in your report to SourceForge:

Browser type
User ID/Nickname or AC
What steps caused this error
Whether or not you know your ISP to be using a proxy or some sort of service that gives you an IP that others are using simultaneously.
How many posts to this form you successfully submitted during the day
* Please choose 'formkeys' for the category!
Thank you.

Crap flood v27.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v27.0

Slashdot requires you to wait 2 minutes between each successful posting of a comment to allow everyone a fair chance at posting a comment.

It's been 1 minute since you last successfully posted a comment

If you this error seems to be incorrect, please provide the following in your report to SourceForge:

Browser type
User ID/Nickname or AC
What steps caused this error
Whether or not you know your ISP to be using a proxy or some sort of service that gives you an IP that others are using simultaneously.
How many posts to this form you successfully submitted during the day
* Please choose 'formkeys' for the category!
Thank you.

It only works for one or two generations

[Anonymous Coward]

Then the compiler will have several generations implementing the same ISA with wildly different underlying architectures and you are back where you started at.

It only helps because its new, not because its different.

Crap flood v28.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v28.0

Slashdot requires you to wait 2 minutes between each successful posting of a comment to allow everyone a fair chance at posting a comment.

It's been 1 minute since you last successfully posted a comment

If you this error seems to be incorrect, please provide the following in your report to SourceForge:

Browser type
User ID/Nickname or AC
What steps caused this error
Whether or not you know your ISP to be using a proxy or some sort of service that gives you an IP that others are using simultaneously.
How many posts to this form you successfully submitted during the day
* Please choose 'formkeys' for the category!
Thank you.

Crap flood v29.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v29.0

Slashdot requires you to wait 2 minutes between each successful posting of a comment to allow everyone a fair chance at posting a comment.

It's been 1 minute since you last successfully posted a comment

If you this error seems to be incorrect, please provide the following in your report to SourceForge:

Browser type
User ID/Nickname or AC
What steps caused this error
Whether or not you know your ISP to be using a proxy or some sort of service that gives you an IP that others are using simultaneously.
How many posts to this form you successfully submitted during the day
* Please choose 'formkeys' for the category!
Thank you.

A Better Topology

[rokicki]

A better topology for the 8-way is something like:

P--P--P

|||

P--P--P

|||

P--P--P

This increases the memory bandwidth available to the four outlying processors, and it also reduces the average path length between processors.

Those inter-processor jumps will accumulate and dramatically increase the memory latency!

Can anyone do better? Assume four of the processors have three links, and four have two links; try to decrease the average path length.

It helps if the topology can be laid out on a board, too!

Re:A Better Topology

[rokicki]

The middle P should be a + (crossover):

P--P--P

|||

P--+--P

|||

P--P--P

Re:A Better Topology

[Anonymous Coward]

There is a lot of broadcast traffic for snooping, thats more efficient in their topology.

Re:A Better Topology

[rokicki]

No; my topology is still better for the snoop traffic because this traffic can be balanced across the links better.

Consider the top two processors in their topology. There is a total of two links connected between these two processors and all other processors. This is a weak bisection and these two links will carry a lot of traffic.

In my topology the worst-case two-processor bisection has three links.

Re:A Better Topology

[pkesel]

In college we played with an 8-way with CPUs set up as a cube. Each was connected to three others. In fact, it was called the HyperCube. It was a very effective setup.

I'm absolutely guessing, but I'd think that having anything but a symmetrical setup for any processor is going to cause engineering headaches, not to mention logical complexity.

Re:A Better Topology

[jbischof]

howabout a hypercube, or mesh, or 2-way switches with shuffle interconnects between them

Re:A Better Topology

[chompz]

A circle with a circut for directing in the center. Operate the traffic cop at high speed and everything is happy.

Re:A Better Topology

[Anonymous Coward]

Oops failed to see the links in the middle, thought the middle processor was only linked 2 ways.

I thought that because there is a slight problem ... the processor can only be linked 2 ways :) It only has 2 interprocessor HT links.

So what if the processor burns?

[Anonymous Coward]

What cant AMD let occur? Have people make propoganda under the guise of being informative?

The only possible time your heatsink would be off is after shipping (not in the lighting strike sense of course, but lets stick to realistic chances). If it fries then you send it back, just like when any of the other of a million things which can go wrong during shipping do go wrong.

At any other time the temperature thing has 0 impact on most people, unless you are one of the people who builds their own PC's ... but then you should know better, and not care because you know better than to turn it on without the heatsink on properly. Of course some people who build their own PC's are morons, but these kinds of people cant be protected against themselves whatever you do.

Re:So what if the processor burns?

[LighthouseJ]

If you actually thought about the article for more than 2 seconds, you'd realize that the article isn't about "would your processor survive the absence of a heatsink", you should have interpreted it to mean "this is how much your processor manufacturer engineers protect their devices". Any cretin can agree that realistically, a heat sink isn't going to just fall off.

To respond to your very first question, it's this thing AMD can't let happen. AMD needs to innovate and make their products even more feature-filled than Intels if they want to compete, if they are going to let their processors overheat and fry themselves, they obviously don't want to put in the effort needed. If you read the tomshardware.com article in the original post, they tested the AthlonXP processor that just came out a few days ago, and if AMD still hasn't written temperature protection in current-model processors, they can say bye to their substantial market share.

Re:So what if the processor burns?

[Anonymous Coward]

I thought about the article, the moment I was beyond the first 2 seconds I realized it was propaganda and damned Tom for printing it and all the little fan boy's for spreading that crap.

Having to compete on utterly useless features does me a disservice, it just means other innovations which are usefull to me have to suffer.

Who? USErs.

[Penguinoflight]

Users will use it. The Hammer is meant to cut off x86/32. Itanium is just a machine that you brag about being in your server farm. The Hammer is sometimes called x86-64, because it's backward-compatible to a 32bit x86. The result? Likely Intel will be forced to make a desktop version of Itanium. Go amd.

Research picks winners

[Anonymous Coward]

It will be interesting to see which processor the major science computing centers pick for their next generation of clusters. The Athelon has been a favored choice that market due to its better performance in float point calculations and a greater number of instructions per clock cycle. This IPC difference shows the dichotomy between the two companies. While AMD has focused on performance, Intel has abandoned this concept altogether to entice the masses with promises of tera hertz processors that couldn't outperform an overc-locked P III.

what they would do

[Dave_bsr]

They would do anything they could. It's war between AOL and MS, and if AOL could bring a new OS or anything out to the Hammer that people would go for, AOL would do it.
Imagine, new AOL cd: "New! 1000 hours free! Comes complete with AOLinux, for your new Hammer system!" something like that. The idea is, if they can fight MS, and Microsoft's MSN internet service, that is a Good Thing for AOL. Both companies see a fight coming, and if AOL/TW have their act together, something like this would be brilliant...
Right now, people have to deal with Windows being dominant in the OS market...80-90% dominant at that. But if people find that the non-MS supported Hammer is that much better than P4's...which you have to admit it probably will be...and almost as cheap...there might be a spot, a crack in the wall as it were.

Re:what they would do

[cymen]

Right now, people have to deal with Windows being dominant in the OS market...80-90% dominant at that. But if people find that the non-MS supported Hammer is that much better than P4's...which you have to admit it probably will be...and almost as cheap...there might be a spot, a crack in the wall as it were.

But how would they even know their system was faster because they can't run any of their old applications (well, they could with wine, but do we really expect AOL users to setup wine). I agree that it would be cool and I think these small problems could be taken care of but I doubt that AOL wants to get into providing an operating system. There would be a *lot* of small problems to take care of...

Re:what they would do

[ackthpt]

but I doubt that AOL wants to get into providing an operating system.

What do you think they are doing with this whole AOL interface?

A few years back I was in a discussion with some guy with blinders on who issued a statement that no home-user would ever need a system with 1 Gig of memory. The old 640k-should-be-enough-for-anyone quote mis(?)attributed to Mr. Gates is dredged up as an example of shortsighted thinking. Same for this fellow, as he had no concept of where sound and video would go, and subsequent demands on memory. Ok, maybe you have a 1.8GHz P4 or a 1.5GHz Athlon smoking through your sound/video/apps/whatever, but, as I've learned over the years, no architecture remains fast for long. Eventually applications come along, which were written off as impractical or impossible before, and tax the resources to the max.

Imagine AOL viewing Hammer-based systems and the thing with enough horsepower to provide some service while Microsoft views it beneath their dignity to do a port of XP. If it draws customers you'll see some real change in the thinking in Redmond. I think the Hammer is another excellent move by AMD, as it's likely to hit the consumer market, perhaps not first, but with a lot of force when it does.

Re:what they would do

[cymen]

What do you think they are doing with this whole AOL interface?

Oh come on... An operating system == AOL interface? I understand where you are going with your idea but I still doubt that AOL is going to want to support an entire operating system.

A few years back I was in a discussion with some guy with blinders on who issued a statement that no home-user would ever need a system with 1 Gig of memory. The old 640k-should-be-enough-for-anyone quote mis(?)attributed to Mr. Gates is dredged up as an example of shortsighted thinking. Same for this fellow, as he had no concept of where sound and video would go, and subsequent demands on memory. Ok, maybe you have a 1.8GHz P4 or a 1.5GHz Athlon smoking through your sound/video/apps/whatever, but, as I've learned over the years, no architecture remains fast for long. Eventually applications come along, which were written off as impractical or impossible before, and tax the resources to the max.

Where are you going with this? I think anyone with half a clue on /. isn't going to disagree that increases in processor speed and memory compacity are a regular event and there will never be "enough" of either.

Imagine AOL viewing Hammer-based systems and the thing with enough horsepower to provide some service while Microsoft views it beneath their dignity to do a port of XP. If it draws customers you'll see some real change in the thinking in Redmond. I think the Hammer is another excellent move by AMD, as it's likely to hit the consumer market, perhaps not first, but with a lot of force when it does.

Windows XP should run just fine in 32 bit mode on the Hammer like linux runs in 32 bit mode on some 64 bit chips. The whole point of Hammer is that it is so backwards compatible compared to Itanium that it won't be a big pain to upgrade for the end user. Anyone have any proof that XP is going to have a hard time running on Hammer? If I were running Microsoft I'd have someone keep up to speed on Hammer. Then if it is released as expected and sells well I'd make sure we support the processor in 64 bit mode. It shouldn't be too hard after all because Microsoft is working on the Itanium support. What advantage does Microsoft get by supporting Hammer right now?

What I don't see is how you think Hammer suddenly makes possible for AOL a number of things that aren't possible today. I think the big break throughs will come with inexpensive and highly accessible bandwidth. The bandwidth will make the difference for AOL - not the cpu speed. In either case high speed CPUs will be here no matter what...

Re:what they would do

[Bert64]

Actually, Windows NT has already supported the 64bit alpha processor, but running in 32bit mode.

Re:what they would do

[Ziviyr]

Oh come on... An operating system == AOL interface?

Why not? They could hack up Linux and put their client on it. BOOM! America On Linux :-)

New instruction for branch?

[steveha]

Reading the discussion of improvements to the branch prediction, I had an idea: might it be useful to add some new branch instructions, which serve as hints to the branch prediction hardware?

Suppose you have a branch on checking the error code returned by a function. That is what the article called a "static" branch: it almost always branches one way, assuming the function rarely fails. The Hammer will try to detect static branches, but might it be useful to let the compiler use different instructions, the static branch instructions, to tell the branch prediction hardware to assume a certain branch is static?

I guess I don't have a good handle on how difficult it is for the branch prediction hardware to sort out static branches vs. the other kind. Would the new instructions help enough to be worth the costs of extra instructions?

steveha

Re:New instruction for branch?

[steveha]

might it be useful to add some new branch instructions[?]

I received an email telling me that the IA64 already has this: you can specify a static branch that is likely to be taken, a static branch that is unlikely to be taken, and dynamic branches in both likely/unlikely flavors.

Also, even on x86, there are some tricks worth doing. The Linux kernel hackers have started using likely() and unlikely() macros around some branches in the kernel source. GCC can arrange the generated code somewhat differently and it will do some good.

steveha

Two problems...

[Pedrito]

OSes and Compilers. If MS doesn't support the hammer architecture in its OSes and compilers, then AMD is screwed. You can talk all you want about "here's a great chance for Linux to hit the desktop." Ain't gonna happen. Look, I love Linux as much as the next guy, but it's not ready for the desktop. The people that run Linux are primarily programmers and geeks.

For a really viable chip, you need the support of the mainstream, and like it or not, that's Microsoft. If they don't support it with their OSes and compilers, then this will be the death of AMD. I'd hate to see that, but those are the facts.

Re:Two problems...

[Anonymous Coward]

I think it's funny that people say "Linux isn't ready for the desktop" without noticing who has the desktop right now: Microsoft. Windows isn't any more ready for the desktop than Linux, but you don't see that stopping anyone, do you?

It doesn't matter if an OS is "ready for the desktop" or has a decent UI or not. If you preload it, people will run it.

Re:Two problems...

[PeterM from Berkeley]

Actually, Hammer doesn't need OS and compiler support from MS. Hammer runs 32-bit code and existing software fast.

It would HELP if MS had their OS and compiler support Hammer's extensions, but even if MS sits on its ass, that huge legacy market will belong to AMD.

And what about the server market? Well, the server market is *much* more accepting of non-MS operating systems.

I do not see lack of MS support as a certain sign of doom for AMD....

surprised no one's mentioned this yet

[CoughDropAddict]

http://www.x86-64.org/

An AMD sponsored web site with the goal of porting free/open-source software to x86-64. Self-serving publicity stunt? Maybe, but it's nice anyway, and certainly more than we can ever expect to see from Intel.

itanium vs hammer on open source OS

[megabeck42]

Honestly - the open source operating system community stands to benefit from Itanium more than Hammer. Kernel extensions for itanium and hammer are already underway, with gcc targets already developed. There are few 64-bit architectures that are being actively developed anymore. Alpha and MIPS are basically EOL, only UltraSparc survives. If Intel has a fast, efficient 64 bit processor, we can expect Linux to immediately benefit, recompile everything and you're gold. Hammer includes a set of 64 bit extensions to ia-32. IA-64, by all reports, is much better designed. With the ability to recompile all code for native ia-64 bytecode, people don't need to bother with ia-32 emulation. x86 should've died more than a decade ago; unfortunately, it seems that in general the worst technology seems to be most fundamentally embraced.

Thoughts on the 64-bit architecture split

[Junks Jerzey]

On the one hand you have Intel, who is trying to move into *completely* new territory, at least as far as breaking with the x86 past. Scary? Very. When Apple transitioned from the 68K to the PowerPC it was rough going for a long time. The PowerPC was much better for native applications, but those took their time in showing up. And it was much, much slower than a real 68K machine when it was emulating older code.

AMD is taking the incremental improvement route, which makes a lot of sense. But can the non-standard x86 extensions--practically a whole new processor in itself--ever be more than a niche? The 3DNow! extensions were more a novelty than anything else. Some drivers used them, most programs didn't. It's difficult as it is to support all the different computers running similar chips without getting into extensions that only work on a certain percentage of them. Is it worth shipping 64-bit Hammer code just for one market segment? It's not just a recompile; it's an entirely separate QA cycle. Thinking about hobbyists: Will they have both Itaniums and Pentiums around for testing?

And then there's the nagging doubt that we're talking about chips that are already so fast that no one cares--except a certain fanboy crowd--so now we're talking about the difference between 10x more speed than I know what to do with and 20x more speed than I know what to do with. Sure, games and some crazy high-end airflow simulation, but this begs the question of "Is it worth overturning the entire PC market just for those two minorities?"

Re:Thoughts on the 64-bit architecture split

[donglekey]

You and all the other idiot whiners who think they have "too much power" need to stop and think about what you are saying. Can you do realtime JPEG2000 encoding of 1920x1080x60fps on your computer? I didn't think so. Can you even do realtime encoding of 320x240 encoding of DivX ? No you cannot. If you could, you would be much better suited for video conferencing at a higher quality. 3D rendering (non-realtime) will never have enough speed for at the very least, the next 20 years, probably more. Some people said the same thing about 486's, pentiums, and everything else until mp3's came about, until divX and mpeg2 (which still use dedicated hardware) came about, until emulation came about, until desktop pubishing came about, until digital video editing came about, until GUI's came about. JPEG2000 creates an image about 1/4th the size of a jpeg with better quality, but it is very slow to decompress and compress. It will hopefully replace jpegs, and then your webpages will load slower. Just because you don't use your computer's power doesn't mean that there aren't other people pushing the envelope with every extra bit of power they get. I am getting tired of answering posts like this just because you can't think more than 6 months down the road.

Re:Thoughts on the 64-bit architecture split

[Junks Jerzey]

You and all the other idiot whiners who think they have "too much power" need to stop and think about what you are saying.

Sigh. I am a software developer. I write applications in Lisp. I make heavy use of graphic arts tools like Corel Draw. I also use 3D modelling packages. What machine do I do all this on? A 333MHz Pentium II that I bought new in 1998.

Do I have _any_ speed complaints at all? None. It is a zippy system. I can recompile the Lisp system I use--which is written in Lisp--in twenty seconds. I also do a lot of work in Delphi and I've never had a perceptible compile time yet (read "for all intents and purposes, compile time is instantaneous"). Corel Draw just zips along. The 3D modeller is more dependent on the video card than anything, so I put in a GeForce 2 and haven't had any--and I mean *any*--issues with speed.

People who talk of using the power of their 1.4 GHz processor don't have a clue. They like to think that they are a power user of some sort, and in all honesty they don't want to hear otherwise.

Re:Thoughts on the 64-bit architecture split

[Namarrgon]

Sorry, I have to agree with donglekey here. I too am a software developer, in the digital content creation market, and yes, I (and my customers) want every CPU cycle I can get my grubby hands on.

I run a dual Xeon 1.6 GHz machine, and it isn't enough. If we could afford a 6-CPU Alpha AXP box, it wouldn't be enough either. My customers use render farms of 100+ CPUs @ 1+ GHz each, and even that still takes days, nay, weeks to render the hundreds of layers of globally-illuminated 3D that they use. Sure, I can compile adequately fast (though a full build of our whole software tree still takes hours), but to test my image processing code on a sequence of 200 MB film-resolution images requires considerable patience.

Just because your needs don't require anything more than last year's gfx card and last decade's CPU does not mean others are happy to sit around and wait for their more complex tasks to complete. More CPU power means more possibilities. That's why we can now produce visual effects like Final Fantasy, Swordfish and SW:TPM instead of Tron and Wargames, to pick examples from just my industry out of hundreds.

In 1980, I read a column in an early computer magazine wondering why people were so keen on the newfangled 16 bit CPUs, with awesomely powerful 32 bit CPUs on the horizon too! He felt that his 4 MHz Z-80 ran his CP/M word processor & spreadsheet quite adequately, thank you very much. Perhaps you too would be happy with that setup for your current line of work?

Re:Wise Intel RE:Do we need to carry on x86?

[ackthpt]

Well, friend, I happen to run C64 and Apple ][ (6510/6502) emulators on my Pentium laptop because there's still some enjoyment in fiddling with stuff on them. Of course there's little new 65xx software out today, and once an architecture like that of native 64 bit Hammer takes hold, expect fewer 32bit apps to come out. The interesting bit of course is the fork in 64 bit style between Intel's path and AMD's path. AMD is poised well, if Hammer is inexpensive, to carve a large chunk of market out of Intel. I'm still waiting for a shoe of Intel's to drop, because you know once the Hammer comes out that P4 just won't be the coolest toy, and nobody is going to sit still for an Itanium grinding slowly through x86 instructions, etc. Heck if that were such welcome idea, half of the world would be running the cheap Alphas (where you had to convert your executables, whee!) Even the ill-fated Amiga computer, with it's x86 bridge card shows you can't put two processors in a machine, sell it for twice as much and expect it to take the world by storm. (I never understood why they actually considered that, myself, I bought an Amiga to run my Amiga stuff, not act like a goofy PC.) If you can run a 64 bit OS with 32bit windows and stuff in it's own little environment window, and provide an easy pipe between, that even Joe Goldenparachute can use with ease, you've got it made.

"Corporate monopolies stopping progress"--How true

[Anonymous Coward]

... and then SRI (Stanford Research Institute) licensed the GUI and mouse to Xerox who pretty much gave it away to Apple only to have it stolen (yet again) by M$FT/[insert other GUI based OS here]. The more I think about what that "troll" said the more he makes sense. The only real innovation now is faster and smaller clones of the origenal computers (im thinking LISA as I don't recall the name of the Xerox machine)!

Act of God?

[SaDan]

You bet! Because only God would use AMD processors.

Everyone knows WinTel is the work of the devil! ;-)

Wise AMD

[Sloppy]

Don't you think that we've hauled along the old 8086/XT baggage long enough? Do we really need a 64-bit 2GHz processor that can still run an MS-DOS 1.0 executable file, or that needs a multi-stage boot loader to crawl its way up the evolutionary ladder from 16-bit to 32-bit to 64-bit "mode", accompanied by a BIOS that has 6 different ways to map a 400G hard-drive into a 1024x16x63 parameter space?

If you're trying to advance technology, no we don't need that.

If you're trying to sell a product and make money, Yes, you definately need that.

Intel and Microsoft have proven it over and over and over: the market does not want progress. The market will only accept incremental evolutionary change.

Somehow Intel has forgotten this, and they are going down the road to technology instead. Meanwhile AMD is going to "out-Intel" them and get all of Intel's customers.

>

I feel that at some point the best thing to do is walk away from the old architecture and make a fresh start with a new one. Commodore did this when they went from the C-64 to the Amiga.

Yes, but you're a damn fool idealist who likes computers and wants to see them run well. You're not trying to sell chips. So while Intel goes off to recreate the marketing success that Commodore had in the 90s, AMD will go off to recreate the marketing success that Intel had in the 90s.

Re:Point of view from a electronic/computing engin

[roca]

> AMD is trying to do in hard and in real time
> what a software (let's say a compiler:)) can do
> much more easily in a very long time.

NO. It is very hard for a compiler to accurately predict what will happen at run time (for example, which loads will hit in the cache and which will miss). It is much easier for the CPU to collect, predict and use this information at run time.

IA64 pushers talk all the time about how smart the compiler "can" be, but they don't actually have any such smart compiler. That is why their performance sucks.

Furthermore compilers are not going to get much smarter in the near future; just because the technology is needed does not mean it will suddenly appear. Compiler researchers aren't stupid and they haven't been sitting on their hands for the last forty years.

> And a CPU can't see more than a few operations
> ahead whereas the compiler can see the WHOLE
> code.

... until the program makes a call into a shared library that was compiled by someone else.

> Just recompile all your software : to run old
> stuff, use old hardware.

Uh huh. So every single time a new chip comes out, Microsoft et al are going to release new compiled versions of all their software. I don't think so.

Re:"Corporate monopolies stopping progress"--How t

[Anonymous Coward]

Mouse != GUI. Stanford had nothing to do with the development of the GUI.

what really matters

[dbrower]

is whether AMD is going to convince anyone to make servers out of Hammers. Who is around? Dell, uh huh. HP/Compaq, right. IBM? Na ga da. Unless some player from left field does something that is seen as compelling, the hammer is going to be as relevant as the F-20 tigershark or the YF-23. It's great technology, but maybe the window has already closed.

-dB

Crap flood v30.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v30.0

Slashdot requires you to wait 2 minutes between each successful posting of a comment to allow everyone a fair chance at posting a comment.

It's been 1 minute since you last successfully posted a comment

If you this error seems to be incorrect, please provide the following in your report to SourceForge:

Browser type
User ID/Nickname or AC
What steps caused this error
Whether or not you know your ISP to be using a proxy or some sort of service that gives you an IP that others are using simultaneously.
How many posts to this form you successfully submitted during the day
* Please choose 'formkeys' for the category!
Thank you.

Crap flood v31.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v31.0

Slashdot requires you to wait 2 minutes between each successful posting of a comment to allow everyone a fair chance at posting a comment.

It's been 1 minute since you last successfully posted a comment

If you this error seems to be incorrect, please provide the following in your report to SourceForge:

Browser type
User ID/Nickname or AC
What steps caused this error
Whether or not you know your ISP to be using a proxy or some sort of service that gives you an IP that others are using simultaneously.
How many posts to this form you successfully submitted during the day
* Please choose 'formkeys' for the category!
Thank you.

Crap flood v32.0

[Anonymous Coward]

Mod me down plz. I need an ip ban, stat. v32.0

Slashdot requires you to wait 2 minutes between each successful posting of a comment to allow everyone a fair chance at posting a comment.

It's been 1 minute since you last successfully posted a comment

If you this error seems to be incorrect, please provide the following in your report to SourceForge:

Browser type
User ID/Nickname or AC
What steps caused this error
Whether or not you know your ISP to be using a proxy or some sort of service that gives you an IP that others are using simultaneously.
How many posts to this form you successfully submitted during the day
* Please choose 'formkeys' for the category!
Thank you.

FUD

[child_of_mercy]

Sounds like classic dominant player FUD to me.

"Hey Guys, We've got nothing good now, but don't go changing to the other folks product, 'cos our next thing [which doesn't exist yet] will be much better, we PROMISE"

You can only judge by whats on the market today, and predict by what's here tomorrow.

Mckinley is the day after tomorrow.

Re:"Corporate monopolies stopping progress"--How t

[LMCBoy]

"Mouse != GUI. Stanford had nothing to do with the development of the GUI."

You're mistaken. Follow the link I gave in my previous post. Doug Engelbart is credited with inventing the world's first GUI as well as the world's first mouse. At Stanford.

Not really comparable

[jquirke]

The Hammer family is in a completely different boat to the IA64. Although it is "64-bit", it is just really an extension to the current x86 architecture, which is getting old in age.

I am strongly familiar with the IA64 and Hammer instruction sets and architectures, as well as the current x86 architectures. The IA64 is a *fresh start* (not like the Pentium, which was extensions the 486, which was extensions to the 386, which was a [large] extension to the 286/8086 and so on), and offers features like

* Predication (more 'smooth' execution)
* Loop pipelining
* ENORMOUS amount of registers
* Heavily used ILP
* Speculation to take care of ambiguous jumps/memory accesses
* A very interesting way of implementing VM

and a hell of a lot more that I can't think of the top of my head. As well as finally discarding segmentation as well.

Although i have a mixed opinion towards IA64 (I mean, its a hell of a lot more work for those writing assembly for it, not to mention compilers and OS's), In the end, the Hammer architecture, while a big step forward, is merely squeezing whatever is left out of the x86 architecture. It really is just doubling the width of the registers and allowing easy implementation of a 64-bit data type, without offering any of the new optimizing features of the IA64. In my opinion, while the Hammer may have the advantage initially, IA64 will eventually dominate.

Software developers too, are unlikely to want to release two binaries (forget the open source argument for now :-) of their software, for _completely_ incompatible architectures.

So in short, they aren't really competing products, nor are they even aimed at the same market. They still however, are both good for what they are.

--JQuirke

NOS=bad

[digitalunity]

No, NOS is bad. It can do some good things, and some bad things. When a NOS/Gas combustion occurs, it happens much more violently than a forced induction combustion. It's harder on the engine. However, you can make huge power gains with NOS because of it's cooling effect. The lower intake charge temps help reduce pre-ignition and the more dangerous detonation.

I've driven a couple high rever's(Integra's at 8250 rpm). They were a kludge to get going, but with the right gearing, hell of a blast to drive. I haven't driven an S2000, but I got a ride in one. Friggin sweet. But, I digress; If you can stay in your powerband, you're doing just fine.

Wanna jump off the line? You'll have to slip the clutch, 'cuz there just aint no ball busting torque.

Re:NOS=bad

[ThatComputerGuy]

You missed my point.. the parent of my post said that intel's "inflated ghz = nos". Not so.

State of the Linux port

[thorsen]

I have been working for SuSE Labs on the X86-64 port for about a year now, and I thought you might be interested in hearing about the state of the port.

Back in march we saw the first printf ("Hello World\n") succeed in the simulator. This is quite a big thing because it needs a working compiler, binutils, glibc and kernel. Since then we have steadily improved the system. By now we're running a full fledged Linux system in the simulator. The system is partly 64 bit and partly 32 bit. We will use the native 32 bit capabilities of the chip to use 32 bit binaries when that makes the most sense (who needs a 64 bit ls when a 32 bit ls does 64 bit filesystems fine).

By now gcc (C and C++ support), binutils, glibc, gdb, the kernel, ncurses, bash, util-linux, vim etc. have all been ported almost completely. And X runs happily in 64 bit too. Now we need the desktop systems, apache, databases etc.

Shameless plug: I'm giving a one-hour talk about Linux on X86-64 at Linux World Frankfurt next tuesday, october 30th. Here I'll show the system running, give an overview of what porting Linux is and describe the new features for Linux that we have implemented.

Bo Thorsen,
SuSE Labs.

Re:State of the Linux port

[Anonymous Coward]

Bo Thorsen wrote:

> By now gcc (C and C++ support),

Is there a fundamental reason why you don't support Objective C, Fortran and Java yet, or
are these left out because of lack of time ?

[ If there are problems with Fortran on the
Hammer, I'd rather hear it *now* than in
a half years time when 3.1 hits the archives ]

Toon Moene, g77 maintainer.

wasted troll Was:Re:And NetBSD already runs on it

[castlan]

Wow, a nonstarter. Who would have thought that the dissapointment of not even seeing an offensive link could be more offensive than actually putting some effort into a troll?

It's a new generation of trolls, I suppose... leave them hanging and wanting more. I'm seriously dissapointed! "coming soon...." ???

at least link to goatse... have some respect!

Itanium isn't about 64bit

[Otis_INF]

It has 64bit support, just because Intel thought it would be great to put it in, but the MAIN point about the Itanium is the EPIC instruction set: move back to simple RISC like instructions and let the compiler do all the math about branchprediction etc etc. F.e.: when you have a program compiled with a good EPIC compiler, you'll have 8 instructions executed PER CLOCK, thus in theory running your program on 8 CPU's at once. It's 64bit too, but that's just a 'nice feature', not the main issue.

Then looking at the hammer: AMD offers 64bit as its main new feature, but keeping the fat x86 instructionset. Nice, but not a product that will survive for at least 10 years from now, resulting in a quick set of bucks fast, but a slow death in the long run...

Re:Itanium isn't about 64bit

[Bluefire]

Itanium is all about EPIC AND 64-bit. It's intended for big servers where you need 64-bits to be competitive.

AMD:s hammer is nextgen x86 with 64-bits added "just for extras and marketing".

>"when you have a program compiled with a good EPIC compiler, you'll have 8 instructions executed PER CLOCK, thus in theory running your program on 8 CPU's at once."

Well, The Athlon has 3 units, so that doesn't make the EPIC like 8 normal CPU:s. (And don't the Itanium have only 4 units, with Mckinley increasing to 6 or something like that?

And FYI, it's extremely hard making good EPIC compilers for general purpose code. Which is the reason for EPIC:s abysmal integer performance.

Sure, I can sympathize with intels wish to kill x86, but why replace it with something even more kludgy?

So What's a generation?

[budgenator]

In the begining it was easy;

1. 4004 [4 bit word]
2. 8008 [8 bit word]
3. 8080 [8 bit word]
4. 8085/8086/8087 and 8088
5. 80186 (Loser)
6. 80286
7. 80386/80387
8. 80486 sx/dx

each step was somewhat a logical progression. Now we have Pentium's pentium II, Pentium III and Pentium IV; mix in MMX in some, and Xeon branchings in the Pentium series. So do we say at least a plain Pentium, or Pentium MMX? How about a Pentium II?

Actualy I think Intel should rot in hell for putting the CPU vectors at the top of memory space at 1 Meg and working down instead of the more logical bottom working up.

as a ps the only reason I have a windows partion is to run one win16 application.

Re:So What's a generation?

[astrophysics]

> 5.80186 (Loser)

Hey, I still have a 80186 that's been one of my best purchases. It's on an old Intel SatisFAXion 14.4 fax modem. It really was able to receive a fax in the background without slowing down the system even on an old 486.

Re:So What's a generation?

[budgenator]

cool, the only other thing I've ever heard it used in is ratheon's Patriot missile. I all ways wondered what new capabilites could be added if they upgraded the msl to a 386, those 80186's must be getting hard to purchase anyways.

Re:So What's a generation?

[athlon02]

Actually in my town I was lucky enough to find a shop that sells older chips & parts...

I got 10 8086's, a 80186 (AMD I think), an 80286... They also have: 8259A (old interrupt chip iirc), 680x0, 6888x, 256KB SIMM chips... it's like a computer collector's paradise :)

There may very well be one where you're at and about say $1/chip isn't too bad!!

Then go buy an Alpha while you still can.

[emil]

Abandoning a user base is an extremely dangerous thing to do.

DEC orphaned a whole platform (MIPS DECStation) with a long stream of broken promises when Alpha was brought out. The seeds of Alpha's destruction were sown the moment of its birth. If DEC had been wise enough to develop an FX!32 for MIPS and an ability to run Ultrix binaries under OSF/1|Digital UNIX|Tru64, then the end of Alpha might have been a very different story indeed.

And now Intel/HP/DEC/Compaq has aspirations of repeating this sad history.

If AMD can deliver on even half of their promises, then Itanium is finished.

Re:Then go buy an Alpha while you still can.

[mikefoley]

Alpha's problems had little to do with the MIPS fiasco. Alpha's problems had ALOT to do with the fact that Intel "acquired" bits of Alpha without permission and Bob Palmer, CEO of DEC at the time, didn't have the balls to go to court to prove it. Instead, he sold out for $700million and "preferred vendor status" from Intel. Well, that all went in the toilet when he sold DEC to Compaq (and now Compaq is being sold to HP) It was a comedy of errors, mismanagement and poor marketing that doomed Alpha. FWIW, I think Hammer has the opportunity to do VERY well and seriously hurt Itanium. I just wish they were announcing who is going to be building 2, 4, & 8 way servers. Hammers battle will be convincing corporate markets that AMD is something more than the computers their kid runs down in the basement with his friends.

Well, yes, Itanium runs x86 at the speed of a P100

[emil]

Are you willing to spend over $3000 for P100 speeds for your x86 code?

Neither is anybody else. The emperor has no clothes.

Re:Point of view from a electronic/computing engin

[vidarh]

And the next time you change the internal structure of your CPU, everyone with binaries optimized for the older structure are screwed unless they recompile...

Did some things right? Z8000 68000

[Anonymous Coward]

"True, variety and anti-monopolistic practices are good things, but Intel got to where they are because they did some things right."

When the IBM PC came along it didn't use an Intel CPU because the 8088 was superior to Zilog's Z8000 or Motorola's 68000. With its hideous segmented memory addressing, everyone knew the Intel processor was the worst of the three by a wide margin. They certainly didn't get that right.

a quick comparison

[john_uy]

itanium -> server -> big bucks
hammer -> desktop -> no bucks

itanium -> big fat heatsink
hammer -> liquid nitrogen coolant

itanium -> hp and compaq
hammer -> asus, gigabyte, msi

itanium -> 64 way
hammer -> 8 way

:-)

Re:a quick comparison

[Bluefire]

itanium -> big fat heatsink
hammer -> liquid nitrogen coolant

Have you seen the stats for itanium ???

Isn't it > 120W for that monster? The other way around in that comparison should be more fitting

Re:a quick comparison

[Anonymous Coward]

Hammer is not 8 way, its 8 way without any gluelogic. If we are going to do it like that then Itanium is 1 way.

Hammer will probably have about the same power requirements as Athlon, so that doesnt stray too far from Itanium.

Wether they can get any wins with the server manufacturers remains to be seen ... probably not, its still a reasonable design though.

Opportunity for Linux/BSD

[beamed]

While Windows wont be able to optimally take advantage of this processor, Linux will.

If only the GCC-developement team could make better progress...

Best answer for the wrong problem.

[leandrod]

So AMD will have the best x86 implementation ever, and 64 bits too... but what we need is to get rid of both x86 in favor of more efficient, smaller, cooler, RISC chips that use less energy and produce less heat, and of binary compatibility requirements by portable, preferrably free, software.

Wise Apple

[SuperGrut]

I think a better comparison would be when Apple went from 68K to PowerPC. They did their backward compatibility through software emulation.

I think it would be best to have some super duper new architecture that was much faster and do the backward compatibilty through software emulation and not hardware emulation.

I just do not think the Itanium is that new architecture. They are getting lousy speeds from this thing. If it was ten times faster out of the gate than a Pentium 4 I would say Intel is doing the right thing. right now I do not see it.

With all of this I see the Hammer being something I am more interested in even though I think the software emulation route would be better.

About 64-bit ISA and Hammer's MP support...

[mr3038]

If x86-64 succeeds would it be possible to get rid of x86-32? Could SSE and SSE2 be used to get rid of x87 entirely? It would be much easier to compile code for a CPU with 16 64-bit integer registers and 16 128-bit FPU/SIMD registers with direct access without stack or similar kludges. After OS and most of the apps we use support x86-64, AMD could sell a "crippled" Hammer that would be missing x86-32 support including x87 - you'd have to emulate x86-32 but hopefully you'd only have to run old 32-bit apps so your performance would be good enough. You would need new bios without requirement for 16-bit or 32-bit instructions to boot with but your CPU could be cheaper. I don't know if Hammer enforces to use of only "new" instructions when in x86-64 mode, but I would hope so. The question is how much burden the unneeded part of ISA is in the end?

By the way, does anybody know if I can run Hammer in 32 and 64 bit modes "simultaneously" so that I have some of my apps fully 64 bit and other legacy apps? Does it hurt my task switching performance seriously that processor is running in different modes with different processes? Will Hammer be faster for 32-bit or 64-bit code? If I don't need 64-bit address space should I compile my code for 32-bit instead for better performance? I would guess that even though 64-bit instructions are a bit harder to execute due to 2x memory requirements the increased register count would balance the things.

According to the article when Hammer is working in MP system each CPU handles part of the memory; should OS be able to send an application to specific processor according to physical memory it has allocated instead of current load of processor for best performance? If so, does any OS currently support this kind of arrangement? How hard it would be to make Linux support this?

Need direct access to the RISC Core

[Tassach]

In case you didn't know it, Athalon processors don't run x86 code natively. They decode the x86 variable-length instructios into an internal constant-length RISC format. The CPU core then executes these RISC instructions.

What I think needs to be done to the next couple generation of Athalons is to allow programs to bypass the x86 decode stage and access the RISC core directly. This will allow the chip to run legacy x86 executables, as well as new RISC executables in a completely transparent manner. After a couple of years, the x86 decoder can be phased out of the primary product line. This would reduce cost significantly, considering that (IIRC) about 20% of the transistor count on the Athalon is dedicated to the x86 decoder.

Re:Need direct access to the RISC Core

[athlon02]

You're still talking about massive porting to that RISC architecture and the world would likely ask "Why AMD's RISC rather than Vendor Y's RISC?"

you missed a rather important point

[brokeninside]

OK, let stop right now : why isn't the code already efficient ? Because the compiler does NOT care about the inner structure of the CPU so the CPU has to do all the real work. By keeping with the "good old architecure", AMD is trying to do in hard and in real time what a software (let's say a compiler:)) can do much more easily in a very long time. And a CPU can't see more than a few operations ahead whereas the compiler can see the WHOLE code.

In other words, the problem you have isn't with the x86 ISA, but rather with current x86 compilers. If compilers can make better code for EPIC, those same optimization techniques can be used to make better compilers for the x86 ISA.

However, EPIC will be much harder to make backwards compatibility for. Since x86-64 will be more general purpose, one will not need to recompile all of their applications for every generation of the ISA.

While I personally find the idea of compiling all apps to be optimized for one's own hardware has much to commend itself, the mass market does not.

More TLB's

[bored]

Little narrow views.. I'm going to complain about Anandtech's treatment of the larger TLB space. I think that the primary reason AMD increased the TLB size is because when running in 64-bit mode the page tables are 4 levels deep. This means that it takes ~2x as long to service a TLB miss. In order to compensate for this extended time there needs to be more or faster TLB's. This is probably the primary reason for the TLB being so high on the list. Increased IPC ha ha ha.. Did anyone accually look at the microarch? Does the 8 extra GP's and resulting lower register spill count mean anything for IPC? Hmmmm. back to computer arch for someone.

Re:WYSE Intel CBM Apple Atari TI et al.

[castlan]

Hey now! Maybe your pseudo 4096 color graphics were pretty snazzy for playing Strip Poker(c), but that doesn't give you license to ignorantly bash the C-64! (I understand it had actual digitized hand colorized photographs of real women, but I wouldn't know. I never had an Amiga, my Commodore only had wholesome games involving violence and fantasy!)

I seem to remember accessing dark blue, light blue, blue and cyan, all at once! If those were the only four colors available, I can understand why there wasn't much Commodore hosted porn. And perhaps it's painful to hear elegantly composed music when all you know is porno-jazz!

Seriously, it wasn't 160 x 200 @ 4, it was _320_ x 200, with 16 simultaneous colors! And a 9 octave range, from .1 Hz to 4KHz. If it weren't for the lack of a subwoofer, soundtracks to some of the better C-64 games sound just like songs I hear playing in clubs today. Perhaps 320 doesn't seem much better than 160, but you gotta take all you can get when you're working with 64 KBytes of RAM, buddy. If SIDs are painful to hear, then why are they still being traded today? Because the're classics, and it took Trent Reznor making .Mp3s for Quake to outdo them. Well, maybe that's a bit of an exaggeration, but I can still think of C-64 tracks that I'd prefer over that dour uninteresting Quake track that Trent did.

Commodores had harddrives available. There just wasn't much point in collecting porn in 16 colors. And the entire base OS could fit into 20K of ROM for instant-on capability. There was a mouse available too, and even a fully featured and compact GUI environment that is still available today, called GEOS. Anything you wanted was available, it just wasn't necessary. There is even a TCP/IP stack available for Commodore hardware if you're interested. But please, I know your type... just don't go surfing for porn.

Oh, and I didn't grumble, I just never moved to Amiga. I grumbled when Amiga finally pulled Mr Snowcone under, losing them all of the profits from the success of Mr. Sno... I mean, the C-64.

PowerPC

[Anonymous Coward]

The Hammer may be the best if you feel like running windows on a workstation, but nothing comes close to IBM's PowerPC architecture for servers. The UltraSparc 3 are not even a match for the upcoming Power4. Why is anyone even caring about intel and AMD duking it out for the lowest rung on the ladder.

Re:WYSE Intel CBM Apple Atari TI et al.

[Ziviyr]

The 16 simultaneous colors is not a matter of total control, at the resolution you state you can only have TWO distinct colors per 8x8 character block. Thats pretty sad.

Feel free to point me to cool SID tunes that don't sound like they were rendered on a piezo element. As I recall it was all classical stuff. Greatest thing I remember was playing Daisy on my disk drive.

Oh, and my copy of GEOS magically broke one day. Those REL files are weird stuff.

Re:WYSE Intel CBM Apple Atari TI et al.

[castlan]

Damn, I was trying to get more of a rise out of you then that! I failed pretty miserably.... oh well.

Don't feel too bad about the Geos thing... I really wouldn't know. Before I actually did much of anything productive, I formatted the GEOS system disk trying to set something up I think. Ah hell, I did the same thing with my Apple IIGS trying to get the damn thing to see 5.25 inch disks. Never did succeed, but I had a nice clean disk where the system disk used to be.

I actually really liked the graphics on the C-64. It seemed like they handled their limited range better than other machines that supposedly had more potential. At least that's how it seemed when I saw games that were available for multiple platforms. Spy vs. spy had two sided disks for the Commodore and Apple systems... the Commodore version was about 20 times more playable. The AD&D gold box series, like pool of radiance, by all logical estimated should have looked better on IBM and Amiga machines, but the Commodore looked somehow less "cartoony" and more enjoyable.

Now I suppose that doesn't make sense. But it seems like Commodore game developers had a much better grasp of the system... maybe the Amiga was just too far ahead of its time? Yeah, I'm grasping here, I'll admit it. But look at Shrek and the Final Fantasy movie that were out not long ago. Final Fantasy was amazing... for a completely rendered movie, it was incredible, at times you could forget that it wasn't live action. But it still wasn't there, little niggling things would ruin the whole effect. Now Shrek while stunning, still wasn't as technically masterful, with a pseudo-live action effect that obviously wasn't trying for reality, but it worked better, because they had a better grasp of what was capable.

Now, please... feel free to regard this post as total bullshit and moderate it as such. I'm a bitter Amiga hater, Commodore should have stuck with what they had success with and let someone else with better management skills take on the admittedly amazing Amiga which ended up marking CBM's gravestone. Or more accurately, Commodore business machines was just another notch on the Amiga's morbid belt.

I wonder if anybody noticed how clever I was substituting WYSE for Wise, and then throwing in the names of a whole bunch of other companies from back in the day.

Or how stupid I was because I couldn't remember any other of the defunct companies that were in the same class as WYSE once were. Well, I did remember Cray. Jesus, I hope somebody moderates this down.

Re:WYSE Intel CBM Apple Atari TI et al.

[Ziviyr]

I'm kinda unhappy about the C64 being axed too, it was quite flexible and probably should have had portable versions out by now. ;-)

Problem is that those fugly Apples with their fugly "OS" and their fugly displays seem to have shoved out any consideration of the C64. God I hated those Apple things. It was next to impossible to find out how to program them too. Mainly they were ugly on all levels though. :-)