Sign in with
Sign up | Sign in
Your question

AMD cheating?

Tags:
Last response: in CPUs
Share
November 10, 2001 10:07:29 AM

i would like starting off by saying i am an AMD fan.

i think we missed something about the new rating system,
we have come used to jumps of 100Mhz of every release of new processor, we all know that 100PR=66Mhz, this means that every time AMD releases a new CPU it only jumps 66Mhz over the previous model.
that translated into a lose factor from the costumer side, in other words we lose 30% value (thioreticly) from expacted performance.

is this another turn on the PR rating system, or just something to think about and consider in the future..

BTW, was looking AliG show while writing this post, he is so funny:) 

<font color=green>
*******
*K.I.S.S*
*(k)eep (I)t (S)imple (S)tupid*
*******
</font color=green>

More about : amd cheating

November 10, 2001 10:11:19 AM

No, the pr is the comparable tbird performance, since the clock per clock speed of the palomino is greater than the tbird, you are not losing anything. The jump from 1500 to 1566 in palomino terms is equal to the jump from 1500 to 1600 in tbird terms. The pr rating itself has flaws that can be argued till the end of time, but I wouldnt say amd is using it to cheat anyone. (especially considering it charges so low for its cou's anyways)

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
November 10, 2001 10:37:00 AM

well you didnt change anything then, they still jump only 66Mhz from previous model, nobody cares if it has a new core.
two chips forward give you only 130Mhz advance then a 200Mhz.
in my book thats cheating.

<font color=green>
*******
*K.I.S.S*
*(k)eep (I)t (S)imple (S)tupid*
*******
</font color=green>
Related resources
November 10, 2001 12:31:20 PM

LoL, except a pr rating is NOT mhz! AMD never claimed it was. It is a measure of how fast a tbird would have to be clocked to offer equalvalent performance. According to AMD.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
November 10, 2001 12:50:53 PM

Each 100 mhz increase for T-bird vs P4.P4 was taking some advance.Now maybe Pal take some advantange at each 100 mghz increase.P4 is way way easyer to increase clock speed.P4 netburst is more scaleable.They will not be able to push very again.Actual P4 can go at a very high clock speed maybe 4 ghz at the end of 2002. vs a maybe 2.3 for thouroughbred if they can really make o.13 micron work.

Wisdom dont come with time
Meilleur chance la prochaine fois
November 10, 2001 12:51:39 PM

i never said it is Mhz, my point is instead of the 100Mhz steps we are costumed to from one model to the next, we gain only 66Mhz.


<font color=green>
*******
*K.I.S.S*
*(k)eep (I)t (S)imple (S)tupid*
*******
</font color=green>
November 10, 2001 12:56:46 PM

lol, what are you talking about? 1.13 to 1.2GHz is 66MHz. 1.33GHz to 1.4GHz is 66MHz. Just because the 1900+ is only a 66MHz jump doesn't mean AMD will always do that. You're not making any sense LoveGuru.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
November 10, 2001 1:06:06 PM

Actually these PR ratings bring <i>some</i> sense to the figures put in front of "MHz"!

Intel has thrashed the concept of MHz ~= Performance.

They have shown that it not necessary to improve performance by increasing MHz.
Now with the i845 SDRAM chipset and more - a integrated graphics chipset for P4, the i845G coming they will prove to the world that it is also possible to screw the performance <i>even after</i> increasing the MHz!

In such circumstances when AMD has their processors performing better than their adversaries running at same physical speeds, why should'nt they rate them equivalent to them?

AMD never claims PR=MHz, they say its a 1533 MHz processor, but it will beat a P4 @ 1800 MHz. read <A HREF="http://athlonxp.amd.com/technicalInformation/benchmarki..." target="_new">this doc</A> and the PDF available on the page that describes the process of deciding the PR rating.

girish

<font color=red>No system is fool-proof. Fools are Ingenious!</font color=red>
November 10, 2001 1:30:45 PM

Juin the Thoroughbred will actually be a 0.13u manufacture.
there are basically two limitations to increasing the speeds.

<b>1.</b> The process technology: its the micron spec that decides the size of the channel on the silicon, smaller the faster. 0.18u is reaching its limit, but still the older Williamette P4 PGA423 <i>is</i> running at 2.0 GHz and is overclockable to a certain extent. A 0.13u switch will allow AMD to increase the MHz greatly, basically micron is a single dimension, but it reduces the die size and the total area covered by the traces by square of the difference! that means shorter paths for signals, lesser physical impedences and resistance and thus faster clock! Plus, lower power consumption and ability to run faster.

<b>2.</b> The Architecture: The Netburst is clearly in lead on this front, but it is not the only limiting factor. Basically what they mean is that if they increase the number of stages to execute a instruction, it will mean lesser work to do for each stage and hence it could work faster, at higher MHz! That is the individual stages will have smaller circuits that could pass on the results to next stage faster. But at the same time, deeper pipelines also means that if a code switch occurs then the processor has to begin all over again. And code switching is second nature to the working of a processor!

Imagine you are standing in a row for movie tickets. You just reach the window and suddenly the clerk shuts it down and you have to move back to the que on the next windows at the last position and start the wait again!

It is one of the primary reasons why the P4 performs so badly despite increase in MHz. It can be said that this massive MHz compensates for the inherent performance loss within the architecture, and hence brings it to some acceptable levels.

So, the Thoroughberd will offer some clock speed increase which may go beyond 2.5~3 GHz until AMD redifines the architecture. In my opinion, the 10 stage pipeline of Athlon is quite okay for this high speed. P4's 20 stages are huge! I guess AMD will settle for say 16 stages in their next architecture. But they will probably continue with the current one even with Barton, 0.13u SOI! Maybe they are confident enough to run it that high at 4.4 GHz as they claim (see <A HREF="http://www.theregister.co.uk" target="_new">http://www.theregister.co.uk&lt;/A>)

girish

<font color=red>No system is fool-proof. Fools are Ingenious!</font color=red>
November 10, 2001 1:35:52 PM

umm..i think i made a small mistake, after the 1Ghz (intel anyway)the jumps became 100Mhz steps, the PR system jumps 100Pts while the Mhz only gains 66Mhz.
this is very misleading.

<font color=green>
*******
*K.I.S.S*
*(k)eep (I)t (S)imple (S)tupid*
*******
</font color=green>
November 10, 2001 2:00:20 PM

um, no it's not. A 66MHz on an Athlon XP is the equavalient to a 100MHz+ jump on an Athlon T-Bird because the Athlon XP can do more work in the same number of clock cycles as the Athlon so your point is not valid. If it was 100+ for every 100MHz then that would be very misleading but 100+ for 66Mhz is not.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
November 10, 2001 3:33:31 PM

we would gain 33 more Mhz for every 100Pts otherwise, i dont see what are you tying to prove?

<font color=green>
*******
*K.I.S.S*
*(k)eep (I)t (S)imple (S)tupid*
*******
</font color=green>
November 10, 2001 3:51:36 PM

I think it has more to do with the FSB multiplier. The XP runs at 133(266), so each step is jumping up at a .5 multiplier.

Previous Athlons ran at 100(200) FSB, with a 3:4 multiplier on the mem (100 FSB, 133 Mem). The few Athlon T-birds that were 133(266) did alternate at 66 Mhz (1.33 to 1.4, ect).

Chesnuts roasting on an open CPU
Bill Gates nipping at your wallet
November 10, 2001 3:53:44 PM

but it make the PR ratings useless. The point AMD is trying to say is that the Athlon XP is faster than the Athlon per clock!

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
November 10, 2001 3:58:41 PM

Quote:
the pr is the comparable tbird performance

No, PR rating is suposedly the comparable rating to a Pentium Northwood at that speed.

Quote:
The benchmarks show quite clearly that AthlonXP 1800+ is able to beat Intel's flagship Pentium 4 2 GHz in the majority of benchmarks. From that point of view, the AthlonXP at 1533 MHz would have rather deserved the model number 1900+ or even 2000+. However, AMD was basing its model rating on the expected performance of the next Pentium 4 core 'Northwood', which will be performing better per clock than current Pentium 4 CPUs.

From <A HREF="http://www6.tomshardware.com/cpu/01q4/011009/athlonxp-1..." target="_new">AMD Introduces AthlonXP Processor - Performance Matters!</A> At Tom's Hardware guide.

Chesnuts roasting on an open CPU
Bill Gates nipping at your wallet
November 10, 2001 4:31:47 PM

Quote:

No, PR rating is suposedly the comparable rating to a Pentium Northwood at that speed.

No, go to AMD's website, they're comparing the T-Bird




AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
November 10, 2001 4:31:48 PM

Quote:

No, PR rating is suposedly the comparable rating to a Pentium Northwood at that speed.

No, go to AMD's website, they're comparing the T-Bird




AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
November 10, 2001 4:40:32 PM

An increase of 66mhz on an XP is equal to a 100mhz jump on a Pentium 4.

It's equal performance. You have to think beyond "Mhz Alone" because if a processor can do more per cycle then another processor, then that processor can have a lower mhz speed and still do equal performance.

XP - 9 instructions per cycle
P4 - 6 instructions per cycle

Do the math. It comes out to XP 66.6_mhz = P4 100mhz.

if(GetSystemMetrics(SM_PROCESSOR) != AMD_PROCESSOR)
{
SendMessage(hwnd,WM_CLOSE,0,0);
}
November 10, 2001 4:44:49 PM

lol, why do you have C++ to close an app when an AMD processor is detected? BTW, is AMD_PROCESSOR a predefined constant or what?

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
Anonymous
a b à CPUs
November 10, 2001 5:23:56 PM

But one question do the regular Joe blow consumers care what the pr rating is or that it compares to an Athlon?

Take me parents for instance they buy a computer what do they call it a Compaq 800 that’s how they and many people identify computers. They don’t know jack about the CPU's inside or care really like. Now take my dad good man pretty smart dude at least when it comes to building houses.

So I went to staples with them for their next computer. Then one of those annoying sales guy was like blah blah 1800+ “im like don’t you mean 1533” guys like well “ya”. My dad pipes in “well why the hell are they saying its a 1800+ when it only goes 1533” guy mentioned the Athlon then my dad like “ok I don’t have a clue what the Athlon is. So why wouldn’t I buy the Athlon they don’t lie about its speed do they?” Sales guy was like no but blah blah blah performance. My dad looked at the gentlemen “oh ok performance eh wow um well…? I buy machines for how fast they really are not how fast they are to some other computer chip.” Then the Sales guy was like “well most people find the benchmarks as to be important factor and these chips win them time and time again.” Then my dads like “ok I don’t you what they are either and could really care less.” Then my dad was getting more annoyed as the guy threw the benches at us I was like ok maybe we should go to Radio Shack and you can buy like a Athlon or something then we left.

Point of that whole real event recap was to kind of show the pr rating just pisses people off and confuses the [-peep-] out of guys, like my dad cause I took about 3 hours of explaining so that he understands how that Athlon can get justified to be a 1800+... he still don’t buy it as a truthful.

-Spuddy


<font color=red>Being Evil Is Good. Cause I Can Be A Prick And Get Away With It.</font color=red> :lol: 
November 10, 2001 6:16:00 PM

AMD is fooling ppl, anybody that thinks the PR is derived from the tbird is a sucker and the real fool!

AMD is gonna look very bad in the upcoming months! they will get rid of the PR system when Intel releases the monster P4 o'clockable to 3.5GHz of true mindnumbing speed.

the 3400+ PR blatant masquerade will be history before it even gets out.


"<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"
November 10, 2001 6:46:59 PM

but when you run a 133/266 FSB you can only go up in jumps of 66Mhz in real clock speed. The P4 scales on the '00 because it is a 100FSB quad pumped (or whatever). P4 will be going up in 66/133 jumps one northwood and pc1066 are up and running...

-* <font color=red> Under Offer </font color=red> *-
email for application details
November 10, 2001 6:50:43 PM

I agree with you LoveGuRu, in fact, I posted this very same thing in the AnandTech forum.
The consensus there seemed to be that
1) It makes no difference to us, since we know how the processors actually perform and,
2) AMD's "true performance initiative" it just another way to dumb down the masses rather than reflect the true performance of the CPU's.
3) It makes no difference to Joe Average, because it's no more inaccurate than the more MHz = more performance philosophy they were buying on before.

At worse AMD’s “rating” system is disappointing, and we already knew that.

"Ignorance is bliss, but I tend to get screwed over."
November 10, 2001 6:58:43 PM

wow, I can't believe all these stories I'm hearing? I don't know anyone that likes the Pentium 4. All my relaties and friends agree that the Athlon is the superior chip. In fact, my friend says he wants an Athlon system and nothing but an Athlon system, he's like anything but an Athlon is slow. I don't know where you live but in my area everyone agrees the Athlon is superior.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
November 10, 2001 7:03:20 PM

Quote:
Q: What do the 1900+, 1800+, 1700+, 1600+, and 1500+ numbers mean?

A: These are model numbers. AMD identifies the AMD Athlon XP processor using model numbers, as opposed to megahertz, such as 1900+, 1800+, 1700+, 1600+ and 1500+ versions. Model numbers are designed to communicate the relative application performance among the various AMD Athlon XP processors, as well as communicate the architectural superiority over existing AMD Athlon processors. The AMD Athlon XP processor 1900+ will outperform an Intel Pentium® 4 processor operating at 1.9GHz on a broad array of end-user applications.

Found at <A HREF="http://www2.amd.com/us-en/Processors/ProductInformation..." target="_new"> AMD's Athlon XP FAQ</A>

Chesnuts roasting on an open CPU
Bill Gates nipping at your wallet
November 10, 2001 7:14:31 PM

A large part of the market is not clued in completely as to the difference between an Intel Chip and an Athlon chip. There was a time, a long time ago, where a non-intel chip ment you were in trouble with a large number of applications. Now, that is not the case, but history and reputation are often what drive the public. Intel is still 70-75% of the market, and it is hard at times to find reputable companies (Dell, Gateway, Compaq, ect) building Athlon systems.

Reputation comes with time, and sometimes comes well after things have changed. It tends to lag behind a bit.

Chesnuts roasting on an open CPU
Bill Gates nipping at your wallet
November 10, 2001 7:17:20 PM

umm..thats a good point, didnt think of that.
still PR is crap, in 1 year or so say 3 new cores forward they will still compare it to the old CPU core?
seems to me like a bad thing coming towards AMD, seems they didnt think that far ahead or they gona do something else.

i just hope they drop PR b4 they make the costumers think too much, we wouldnt want that would we..

fotunatly i do understand and know the real life performance of each CPU, i dont need no PR system to tell me whats better.
the simple joe can buy an XT for all i care..
btw, Mhz do represent one of the real performance factors, yet PR rating represent really nothing, if for say they will have a CPU that would gain a 2:1 ratio over the old core it would have to jump 250Pts every 66 Mhz, this will make even me somewhat comfused and annoyed.

<font color=green>
*******
*K.I.S.S*
*(k)eep (I)t (S)imple (S)tupid*
*******
</font color=green>
November 10, 2001 7:30:08 PM

Why do care about AMD, Intel or that average Joe Blow? If Joe Blow is stupid enought to get a 2GHz P4 with 512MB SDRAM for $3000USD then let him, it's not our problem. It's enough to know that you know what you're doing and you know you're doing the right thing.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
November 10, 2001 7:50:41 PM

Guru, There was a T-Bird 1.33ghz, which I guess you must have forgotten about. The speed increases have never been uniformly 100mhz speed increases. I really don't think your concern holds much merit. The overall price/performance is far more important in the end.

When all else fails, throw your computer out the window!!!
November 10, 2001 8:00:25 PM

See Bront? You found the answer right in the FAQ!

What does it say? Broken down into the pertinent parts it says FIRST:

"Model numbers are designed to communicate the relative application performance among the various AMD Athlon XP processors, as well as communicate the architectural superiority over existing AMD Athlon processors. "

This is the meaning behind the model number. It then continues to mention the following:

"The AMD Athlon XP processor 1900+ will outperform an Intel Pentium® 4 processor operating at 1.9GHz on a broad array of end-user applications. "

This last is just an aside, so to speak, or a statement of generally known facts from benchmarks. The actual answer is the previous one though.

Mark-

When all else fails, throw your computer out the window!!!
November 10, 2001 9:36:05 PM

This is the dumbest argument I've ever seen.

The PR rating is based on performance not on Mhz. What AMD is trying to say is that the P4 needs to increase their Mhz much more to get the same performance. The Mhz isn't the only performance benchmark anymore. If you think you're being cheated out of Mhz, fine. Who cares? You aren't getting cheated out of performance. Look at the benchmarks before making these types of arguments.

If we're talking about the average joe, then they should also make an educated decision before buying a system. They need to look at the benchmarks. Unfortionately they don't so AMD needs this stupid system to level the playing field to the point where all the numbers match up performance wise between them and Intel. If you ask me I think the PR numbers are a bit low.

<font color=red>God</font color=red> <font color=blue>Bless</font color=blue> <font color=red>America!</font color=red><P ID="edit"><FONT SIZE=-1><EM>Edited by dhlucke on 11/10/01 03:42 PM.</EM></FONT></P>
November 10, 2001 10:29:52 PM

You must of misread the logic of the if statement.

If the processor is != (not equal) to an AMD, then it closes. There is no System Metric of SM_PROCESSOR that I'm aware of, and AMD_PROCESSOR is made up as well. It's simply displaying my alliegance to Advanced Micro Devices :) 

if(GetSystemMetrics(SM_PROCESSOR) != AMD_PROCESSOR)
{
SendMessage(hwnd,WM_CLOSE,0,0);
}
November 10, 2001 10:38:41 PM

ohhhh, I didn't see the !. I haven't used C++ in a while (I use Delphi now) so that's why I missed it. lol, that's a nice one.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
November 11, 2001 12:19:40 AM

i hpe that you are better at love than you are at this pr rating

:cool: :eek:  :redface: :frown: :lol:  :mad:  :eek:  :smile: :tongue: :wink:
November 11, 2001 1:16:28 AM

So, the Thoroughberd will offer some clock speed increase which may go beyond 2.5~3 GHz until AMD redifines the architecture. In my opinion, the 10 stage pipeline of Athlon is quite okay for this high speed. P4's 20 stages are huge! I guess AMD will settle for say 16 stages in their next architecture. But they will probably continue with the current one even with Barton, 0.13u SOI! Maybe they are confident enough to run it that high at 4.4 GHz as they claim (see http://www.theregister.co.uk)


Personal opinion and AMD marketing.o a architecture that can scale 500 mgz to 4.4 ghz.Get real.Some time you hit the max.

Wisdom dont come with time
Meilleur chance la prochaine fois
November 11, 2001 1:19:17 AM

From AMD Introduces AthlonXP Processor - Performance Matters! At Tom's Hardware guide

Hey i was the 1 to say that.They have say that way after me.

Wisdom dont come with time
Meilleur chance la prochaine fois
November 11, 2001 1:22:32 AM

An increase of 66mhz on an XP is equal to a 100mhz jump on a Pentium 4.

It's equal performance. You have to think beyond "Mhz Alone" because if a processor can do more per cycle then another processor, then that processor can have a lower mhz speed and still do equal performance.

XP - 9 instructions per cycle
P4 - 6 instructions per cycle

Do the math. It comes out to XP 66.6_mhz = P4 100mhz.


That was not the case with T-bird.Xp is not 33% faster that a T-bird.



Wisdom dont come with time
Meilleur chance la prochaine fois
November 11, 2001 2:23:40 AM

Ok.

I was under the impression that the PR rating is comparing the XP processor to the Pentium 4. Not the XP processor to the T-Bird. When you look at the benchmarks at AMD's website, they are comparing the XP and the P4.

if(GetSystemMetrics(SM_PROCESSOR) != AMD_PROCESSOR)
{
SendMessage(hwnd,WM_CLOSE,0,0);
}
November 11, 2001 2:34:47 AM

Heh heh...

Now, if you really wanted to sizzle someone's sausage, you should go around to all the pentium 4 2.0 gHz and slap 1800 PR rating stickers on them.

Customer: So it's a 1.8gHz chip?
Sales: No, it's 2 gHz... but it runs as fast as an AMD 1.8 gHz processor.
Customer: Oh... so I want the AMD 1800 then!
Sales: No.. you want the AMD 2000.
Customer: Huh? That's 2 gHz isn't it? I only want 1.8.
Sales: No, the AMD 2000 runs at 1.8 gHz. 2000 is it's PR rating.
Customer: So, it runs as fast as a pentium 2000?
Sales: Actually it runs as fast as a 2.1 gHz pentium.
Customer: That's the pentium 2100?
Sales: No... the pentium 1900.
Customer: That's only 1.9 gHz!
Sales: No, 2.1 gHz... but PR rating is 1.9
Customer: Bugger it...
Sales: You could always go Cyrix...


How long is a piece of string? The same shape as something beige.
November 11, 2001 3:47:29 AM

Bront, the performance rating is derived from the old tbird performance, but they are selling against the p4, a 1900+ xp will punk a 1.9ghz p4, but the 1900 comes from the tbirds performance none the less.


Thanks to everyone who pointed this out before I got back to take care of it.

::closes all of the browser windows with evidence for the pr=tbird debate::

Actually


http://www.theregister.co.uk/content/archive/22145.html
http://www.theregister.co.uk/content/3/22134.html
http://www.theregister.co.uk/content/3/22047.html
http://www.theregister.co.uk/content/3/22228.html


The site thinks that the pr rating is in fact versus the p4, but acknoledges that amd claims otherwise. Whatever you may believe the pr ratings are designed to do, what amd claims is all that matters, and amd claims it to be related to tbird performance.


That said, increasing the clock speed by 66, and the pr by 100 is not cheating as they maintain their pr system and it is independantly verified by impartial sources.(the company they got to verify the pr system, I forget who it was though)



~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
November 11, 2001 4:11:05 AM

The benchmarks were done by Arthur Anderson Consultants. Here is the <A HREF="http://athlonxp.amd.com/technicalInformation/benchmarki..." target="_new">doc</A> that will describe everything in detail.

girish

<font color=red>No system is fool-proof. Fools are Ingenious!</font color=red>
November 11, 2001 3:20:16 PM

"So, the Thoroughberd will offer some clock speed increase which may go beyond 2.5~3 GHz until AMD redifines the architecture. In my opinion, the 10 stage pipeline of Athlon is quite okay for this high speed. P4's 20 stages are huge! I guess AMD will settle for say 16 stages in their next architecture. But they will probably continue with the current one even with Barton, 0.13u SOI! Maybe they are confident enough to run it that high at 4.4 GHz as they claim (see http://www.theregister.co.uk)"

Athlon is not designed to scale to 4.4ghz. Hammer is planned to scale to PR4400 by end of 2003. Athlon will scale to PR2600 by around mid 2002. Beyond that is up in the air according to AMD.

As for next gen pipes.. Hammer will have 12 or 13 compared to Athlons 10. Hammer is still keeping relatively short pipeline compared to P4, probably to ensure optimal performance per cycle.

Mark-

When all else fails, throw your computer out the window!!!
November 11, 2001 3:48:03 PM

You're forgetting the Hammer is going to be a 64-bit ready processor, so in optimized apps, it can be over twice as fast as a P4.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
November 11, 2001 3:50:44 PM

and PR4400 is equal to what? compared to what? ipcs: how many per clock tick? what is the true speed? 2.0GHz, 3.0GHz surely not 4.4GHz?



"<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"
November 11, 2001 4:38:35 PM

Probably PR4400 compared to the T-Bird. I'm guessing it'll be a little over 2GHz.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
Anonymous
a b à CPUs
November 11, 2001 4:42:27 PM

>what is the true speed?

Who really cares ? We are not making fun of the Itanium running at 800 Mhz, are we ? Its performance that matters. In some apps (well, just a few, maybe one.. the SSL thing) , the Itanium runs circles around the P4/athlon/Mips/etc. If I am buying a server platform, thats what matter, not the Megahertz, or MB cache, or any other spec. If Hammer is as fast as I would like it to be, I couldnt care less what frequency it runs. If it outperforms every other cpu at half the frequency, fine.. I want one then.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
November 12, 2001 1:34:28 PM

Unfortunately, for Intel, right now, Itanium apparently has a bug in the faster speed chips preventing their shipment now. This according to Compaq. So, the fastest speed Itanium currently is 700mhz I think.

Mark-

When all else fails, throw your computer out the window!!!
November 12, 2001 2:16:59 PM

well, true the Athlon is not scalable for that kind of speeds. but there is <b>one more thing that governs the frequency scaling</b> of a processor.

as i said in my previous post, the fabrication technology, if it could get the components smaller and faster the individual stages become obviously faster allowing the processor in totality to run faster.

a switch to 0.13u from 0.18u could allow the AthlonXP to scale almost double if not more. maybe AMD wont make a AthlonXP 3066 MHz ~ model 4000+ but overclockers might be able to get the speed. maybe wishful thinking, but it could be. we will have to wait for a couple of years to know.

the 4.4 GHz Hammer chip (its supposed to be real and not <b>IR</b> - Itanium Relative) will be later in 2003 and will be a 0.09u fab. and it is supposed to be the begining! imagine what 0.09u would make a processor MHz scale to? i wish AMD manufacture Athlons with 0.09u. I cannot afford a Hammer! ;-)

girish

<font color=red>No system is fool-proof. Fools are Ingenious!</font color=red>
November 12, 2001 3:08:00 PM

Yes, it does make sense when you break it apart. Sorry for the mistake. Interesting that the numbers do seem to have a dual meaning though.

Chesnuts roasting on an open CPU
Bill Gates nipping at your wallet
November 12, 2001 3:08:00 PM

Yes, it does make sense when you break it apart. Sorry for the mistake. Interesting that the numbers do seem to have a dual meaning though.

Chesnuts roasting on an open CPU
Bill Gates nipping at your wallet
!