Sign in with
Sign up | Sign in
Your question

very cool article about the itanium and hammer

Last response: in CPUs
Share
April 24, 2002 1:56:40 PM

becareful readng this article! It has lots of computer science terminology. Have fun! :) 

Itanic: It's all academic now (Official)
By Andrew Orlowski in San Francisco
Posted: 04/23/2002 at 02:37 EST


The name Itanic, coined here several years ago by Mike Magee, for Intel's IA-64 processor has been formally adopted by academia.

Nick Weaver, a 28-year old graduate student and researcher, teaches computer science classes at the University of California's Berkeley school, and as you can see from his "special topics" class, week 16 next month will be devoted to the "Voyage of the Itanic".

"Itanic describes the architecture very well," he tells us, explaining that the processor contains great ideas and "beautiful features" that ultimately were compromised by terrible subsequent design choices and "feature creep". We invited him to elaborate:

"The first good idea was one they explored several years ago in a paper from HP. That explains that if you have 64 or 28 registers with 64 or 128 condition registers and every instruction is being conditionally executed, then to get to a statically-scheduled superscalar - you'd get benefits of VLIW- issue logic, which is very simple, without the upgrade problem that VLIW has."

What upgrade problem?

"VLIW doesn't scale because the compiler statically issues for the number of function units in the VLIW architecture . So for a new version of VLIW you have to recompile. Transmeta gets around this by always recompiling, and it's not a problem in the DSP community. But it is a problem if you want to do a 'general purpose' processor"

Nick commends deferred exception handling and low-cost checkpointing as two "beautiful features" of the IA-64 architecture.

"This also interacts well with the speculative techniques derived from the first part: You can speculatively execute both sides of a branch, allow cache misses to be errors which are deferred, and you only take the penalty if there is both a cache miss which is on the branch who's result you want. They make this very nice by propagating error
conditions.

Icebergs
But then the trouble began. His lecture calls Itanic an "unquestionable disaster..."greatly increasing implementation complexity without really providing a benefit to either compiler writers or to performance".

Like what?

"The rotating register file, and the register window notion - they have a thousand registers in there, and the larger the register file the slower it is and the more it costs.

"The rotating register file is there to make software pipelining easier - fills - but it's not actually a big win - and if you ask the compiler writers they go 'why do you this?'". The register window also incurs a performance penalty when doing garbage collection, he suggests.

"The other issue is that combining both features requires that the processor be able to effectively arbitrarily remap a number of logical registers - 128 to the physical registers, which number arounda a thousand. This occurs in a traditional out-of-order machine, but the whole point of EPIC is was to enable lots of parallelism without introducing these complications."

As for the future, he says it isn't completely hopeless. " Once you add a feature to the instruction set it's hard to take it away," he tells us.

Itanic could be helped by a process shrink and wider issue width, by real vector (Cray-like) instructions, but most of all by compiler improvements.

"The right compiler can theoretically allow Itanium to issue 6 instructions/cycle, but at the same time, if the compiler speculates too heavily, this wastes too much effort and performance suffers."

"The MMX and SSD [Screaming Sindy] instruction sets were attempts to build a vector ISA, but they have a few issues, notably too small a vector. Vector machines have a bit of a bad reputation today, one that is, in my opinion unjustified.

Yammer the Hammer
No, he hasn't seen what he describes as "rumors" of Yamhill, Intel's own 64bit skunkworks project, but commends AMD's Hammer as a better approach.

"I think AMD is on the right track," says Nick.

"They've made the core simpler, and that makes it smaller, leaving room for much larger caches."

"The Hammer approach is 'we know how to do a CISC to RISC, how to make that RISC very fast, we know what few changes in the instruction set architecture would make it lots better - 16 general purpose registers, 16 floating point registers instead of the 8 entry stack from the 386 days - so let's do that'.

"This also gave them the ability to really concentrate on the interfaces. The memory interface is on the processor, and the traditional bus has been replaced by networks to communicate with the I/O. This allows glueless 4 way SMP setups, better I/O bandwidth, and better memory latency. It cuts out the chipset when going to memory, which saves 2 pin crossings and a bunch of traditionally slower chipset logic."

"If you want backward compatibility and performance, go Hammer," he recommends. "If you want backward compatibility and performance isn't such an issue, buy Transmeta to translate that old code."

Just my opinion, says Nick. Anyone beg to differ? ®
http://www.theregus.com/content/23/24726.html

<A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>
April 24, 2002 10:31:41 PM

what do you say exactly, its a nonsense.
there is just one word in that sentence.
rofl.


<i>if <b>you know</b> <font color=white>you don't know<font color=black>, the way could be more easy ...<font color=red>
Related resources
April 24, 2002 10:36:13 PM

<i>the frog which would like to be biggest than the buff...</i> lol


<i>if <b>you know</b> <font color=white>you don't know<font color=black>, the way could be more easy ...<font color=red>
April 24, 2002 10:38:33 PM

Out of everyone on this forum, I don't think you're exactly qualified to correct his grammar.

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
April 24, 2002 11:05:49 PM

its not my fault if you didnt caught the meaning of this one. rofl

<b>NOW, YOU CANT STOP THE ALIEN MACHINE, ITS TOO LATE...</b>

btw, teaching is not a vocation but an occupation.

<i>& please man, dont get this one yet again word by word. lol</i>



<i>if <b>you know</b> <font color=white>you don't know<font color=black>, the way could be more easy ...<font color=red>
April 24, 2002 11:08:41 PM

If you can't say it clear enough to be understood, then it <i>is</i> your fault. If I'm the only one who apparently misunderstood, then it would be my fault.

And vocation and occupation are the same thing.

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
April 25, 2002 1:07:19 PM

here is the truth.
noway! they havent the same meaning.


<i>if <b>you know</b> <font color=white>you don't know<font color=black>, the way could be more easy ...<font color=red>
April 25, 2002 1:26:53 PM

Bump is used on the anandtech forum and others as well... they do that to bump the post back to the front page. I bumped it back to the beginning of the page so people can read it as it is very interesting.

It is very common throughout the forums'.

<A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>
Anonymous
a b à CPUs
April 25, 2002 1:29:00 PM

Actually (and back to topic), I thought the article was pretty good myself.

The new article on AMD's Opteron (new SledgeHammer name) was interesting. I only hope AMD can work out the bugs (ie, latency issues) on all the registers and interfaces.

Their 2nd issue will likely be heat dissipation (from looking at that first-silicon case). If it can't fit in a standard 19-inch rack mount in a 1U or 2U case without heat problems, it will be of little use competing with Blade servers.


AMD K6-2@500MHz... takes a lickin' & keeps on tickin'!
April 25, 2002 1:49:33 PM

Yes, back on topic ...

Giving that this is AMD's very first own product that is not intels, i would say that the Opteron will be ironed out. I think the 64-bit athlon is what will be a little buggy as it is the first to come out. After that then the Opteron will be released. Giving AMD time to iron out the bugs for the Opteron and first impressions on the server market. Clearly that is AMD's goal right now. To enter the server market.

I think AMD's 2nd generation of the 64-bit athlon will be something fantastic. Unless AMD provides it's own chipsets AMD's Athlon 64 will be a little buggy in my opinion at first.

Heat is as usual a big issue. I think that is why the Opteron is so much bigger. To spread the heat maybe? Just theories.

<A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>
Anonymous
a b à CPUs
April 25, 2002 2:49:26 PM

Bump.
Anonymous
a b à CPUs
April 25, 2002 2:50:47 PM

Bump(2).
April 25, 2002 4:03:49 PM

Quote:
vo·ca·tion Pronunciation Key (v-kshn)
n.
A regular occupation, especially one for which a person is particularly suited or qualified.


Bite me

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
April 25, 2002 4:10:47 PM

Quote:
Bite me

At least you didn't say 'Byte me'. You never know who might have tried to send corrupt packets to your IP address then... ;) 

<pre>Join PETT.(People for Equal Treatment of Trolls)
Trolls:Keeping bridges clean 'n safe.</pre><p>
April 25, 2002 4:20:32 PM

Quote:
Giving that this is AMD's very first own product that is not intels

<b>???</b>

Quote:
I think the 64-bit athlon is what will be a little buggy as it is the first to come out. After that then the Opteron will be released. Giving AMD time to iron out the bugs for the Opteron and first impressions on the server market.

So you are saying that AMD will be using the average consumer as a guinea pig for their server goals?

Quote:
Unless AMD provides it's own chipsets AMD's Athlon 64 will be a little buggy in my opinion at first.

I am without doubt on this. The new CPU is so different that it is bound to require a very good reference platform from AMD for third party vendors to make their own chipsets work right. That is, if third party vendors are even inspired to make chipsets for it.

Quote:
Heat is as usual a big issue. I think that is why the Opteron is so much bigger. To spread the heat maybe? Just theories.

Cute theory, but it probably has a lot more to do with the x86-64 extensions, the HyperTransport, the memory controller, the bigger cache, the improved pre-fetching, etc., etc. There is a lot of new stuff going into the Opteron, and this stuff has to reside somewhere. Hence a bigger chip.

<pre>Join PETT.(People for Equal Treatment of Trolls)
Trolls:Keeping bridges clean 'n safe.</pre><p>
April 25, 2002 4:42:52 PM

Quote:
???


I believe he meant it's not just a clone of Intel's products, but a new design. This is debatable, of course.

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
Anonymous
a b à CPUs
April 25, 2002 5:03:28 PM

yes but this is not a human disease...
April 25, 2002 5:03:42 PM

I'm only quoting from Jerry Sanders on what he said. Makes sense too ... The Hammer architecture is completely AMD's. Talking about technology.

"So you are saying that AMD will be using the average consumer as a guinea pig for their server goals?"

NO! I SAID: "I think the 64-bit athlon is what will be a little buggy as it is the first to come out. After that then the Opteron will be released. Giving AMD time to iron out the bugs for the Opteron and first impressions on the server market."

Don't spin it around saying i said something entirely different! If you can't understand what i said then say so and i'll reword it. You cannot enter my mind and assume i ment this or that. I suggest you practice NOT spinning peoples words around.

This is very typical in the hardware business. EVERYTHING that is new and to first come out is always buggy. It depends on how strict the rules are too. It is common for the first batch of whatever hardware it is that is new to be somewhat buggy but not severe enough to warrant a recall. Happens all the time. Nothing is perfect on the get go. Why do you think there are "REVISION #'S"? huh? This is common knowledge if you think about it logicly. Just like programming. When a software program comes out it is more then likely buggy. But not buggy enough to warrant a recall. That is why software also has "VERSION #'s", which is the same idea as "REVISION #'s".

And finally there is nothing cute about theory or myself.

<A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>
April 25, 2002 5:40:19 PM

Quote:
This is very typical in the hardware business. EVERYTHING that is new and to first come out is always buggy. It depends on how strict the rules are too. It is common for the first batch of whatever hardware it is that is new to be somewhat buggy but not severe enough to warrant a recall. Happens all the time. Nothing is perfect on the get go. Why do you think there are "REVISION #'S"? huh? This is common knowledge if you think about it logicly. Just like programming. When a software program comes out it is more then likely buggy. But not buggy enough to warrant a recall. That is why software also has "VERSION #'s", which is the same idea as "REVISION #'s".

This is all true and common knowledge. <i>However</i> to specify that the release is purposely being done to be "Giving AMD time to iron out the bugs for the Opteron and first impressions on the server market." is a <i>very</i> different implication than just expecting the first versions to have the standard bugs.

Further, I never <i>once</i> spun your words around. I specifically phrased it <b>as a question</b> to allow you to clarify, affirm, or deny. <i>I</i> suggest that <i>you</i> spend a little more time thinking and a little less time jumping to conclusions. You will look a lot less like a fool that way.

Quote:
And finally there is nothing cute about theory or myself.

I whole-heartedly agree. Perhaps you are familiar with the phrase 'sarcasm'? Then again, perhaps <i>you</i> aren't...

<pre>Join PETT.(People for Equal Treatment of Trolls)
Trolls:Keeping bridges clean 'n safe.</pre><p>
April 25, 2002 5:46:12 PM

Quote:
I believe he meant it's not just a clone of Intel's products, but a new design. This is debatable, of course.

I agree. It is entirely debatable. In fact, considering just how different the Athlon core was from the P3, it is so debatable that to me, it isn't even debatable at all. :)  (I know that hardly makes any sense, but then, that is the point. Claiming that the Athlon was a P2/3 core clone just because it was a flavor of x86 is entirely nonsensical.)

<pre>Join PETT.(People for Equal Treatment of Trolls)
Trolls:Keeping bridges clean 'n safe.</pre><p>
April 26, 2002 12:03:43 AM

Yeah, Great the Dictionary.

from mine:
Quote:
vocation

aptitude for teaching.

Quote:
occupation

activity, action, spryness.


Byte me.


<i>if <b>you know</b> <font color=white>you don't know<font color=black>, the way could be more easy ...<font color=red>
April 26, 2002 1:06:28 AM

your forgetting about the chipsets too ... but you are right AMD has plenty of time to iron out the cpu. I'm more concerned about the chipsets then anything else.

if you are being sarcastic clearly label that you are being sarcastic.

<A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>
April 26, 2002 2:25:56 AM

Is it just me or has this last week been very coincodental???

at the start of the week we heard that the CEO of AMD, Sanders, sell his soul and testify that Microshaft was really a nice company and never did anything bad like be a monopoly and bully other companies.

then, barely a couple of days later AMD gets microsoft approval for a compatible OS.


Despite appearances im not Phsysic. I may need your system specifications to solve your problem!
April 26, 2002 3:04:20 AM

I used Dictionary.com, what'd you use? A French-English dictionary?

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
April 26, 2002 12:10:19 PM

you use <A HREF="http://www.dictionary.com" target="_new">Dictionary.com</A>, i have eyed off.

you jest isnt it ?


<i>if <b>you know</b> <font color=white>you don't know<font color=black>, the way could be more easy ...<font color=red>
April 26, 2002 12:24:52 PM

btw, you "forget" something from YOUR dictionary:

Quote:
4 entries found for occupation.

oc·cu·pa·tion Pronunciation Key (ky-pshn)
n.

1a - An activity that serves as one's regular source of livelihood; a vocation.
1b - An activity engaged in especially as a means of passing time; an avocation.

2a - The act or process of holding or possessing a place.
2b - The state of being held or possessed.

3a - Invasion, conquest, and control of a nation or territory by foreign armed forces.
3b - The military government exercising control over an occupied nation or territory.

& in particular the 1b point which is
Quote:
1b - An activity engaged in especially as a means of passing time; an avocation.

<i>vocation & avocation havent the same meaning.</i>

Byte me.


<i>if <b>you know</b> <font color=white>you don't know<font color=black>, the way could be more easy ...<font color=red>
April 26, 2002 1:02:47 PM

i highly doubt amd is going to release the hammers and have consumers be "beta testers". what other time have they done this? but then again, some companies have the luxury of using consumers as "beta testers" on 2 different sockets and 3 revs. to get the cpu they want.

[insert philosophical statement here]
April 26, 2002 4:24:12 PM

Quote:
av·o·ca·tion Pronunciation Key (v-kshn)
n.
An activity taken up in addition to one's regular work or profession, usually for enjoyment; a hobby.
One's regular work or profession.
Archaic. A distraction or diversion.


Quote:
vo·ca·tion Pronunciation Key (v-kshn)
n.
A regular occupation, especially one for which a person is particularly suited or qualified.
An inclination, as if in response to a summons, to undertake a certain kind of work, especially a religious career; a calling.


That's odd, they seem pretty damn similar to me.

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
!