Pentium V details

[olefarte]

From The Inquirer comes some details on the upcoming Pentium V cpu.

The chip will sample internally at Intel in January 2004 and will take between four to six months to get to market. The Pentium 6 will follow a very similar schedule.

The Pentium V is likely to fly along at between 5GHz to 7GHz, have 2MB plus of level two cache, be built on a 90 nanometer process, and have a stackable design.

The idea seems to be that people can buy a 32-bit module, and then add in the 64-bit processor.

 

[StormBringer]

I read somewhere a while back that the P4 would be the last Intel CPU to bear the Pentuim name. Can't they come up with anything more creative than adding a number. Seems it would have more of an impact that way....

 

[Rick]

Pentum
Pentium Pro
Pentium II
Pentium III
Pentium IV

Athlon
Athlon XP
Athlon MP
Athlon 64

.. Intel AND AMD could use some tips on creative naming alright. I guess AMD has been doing a better job though.. Their names seem to be more "descriptive".

 

[Mictlantecuhtli]

What's in a name? That which we call a rose by any other name would smell as sweet.

--William Shakespeare

 

[LNCPapa]

I think AMD is doing much worse by tricking people with a screwy numbering system.

 

[acidosmosis]

How is it tricking? Intel is the one doing the tricking. I mean, AMD is compensating for Intel's BS marketing by using a number that will show how well the processor performs as compared to Intel chips. It makes things MUCH less confusing. Consumers do buy things because of these numbers and without AMD using the p-rating then they would lose money because consumers wouldn't buy processors due to thinking "oh, this AMD processor is 2.08GHz, but this Intel processor is 2.8GHz" (just an example). And that would not be cool at all. At least with the p-rating consumers get a better idea of the actual performance of these processors.

They were pretty much forced to use them because their processors bear a lower MHz rating than Intel's equivelant's, but that in no way means they perform any worse.

 

[agissi]

Gosh olefarte.. My friend just asked me tonight if 2.8GHz was fast, and im like "oh yea!". With new chips comming in at 5-7GHz, these "fast" 2.8, 3GHz chips are going to be outdated super quick.

Btw acid im all for ya (not that I dont like you LNC )

 

[young&wild]

4000MHZ FSB bus? Some parts of the article is just pure speculation.

 

[Mictlantecuhtli]

Originally posted by acidosmosis
How is it tricking? Intel is the one doing the tricking. I mean, AMD is compensating for Intel's BS marketing by using a number that will show how well the processor performs as compared to Intel chips. It makes things MUCH less confusing.

Not when you're trying to solve problems. Just search how many times people have asked what MHz their CPUs are supposed to run at - you've asked that yourself, too.

 

[MrGaribaldi]

Originally posted by acidosmosis
Intel is the one doing the tricking. I mean, AMD is compensating for Intel's BS marketing by using a number that will show how well the processor performs as compared to Intel chips.

Let's just clear up one thing here... Intel isn't tricking anyone when they tell you what Mhz their cpu's are running at. MegaHerz is MegaHerz no matter how you look at it...

It's not like the Mega-/Mibi- byte discussion, where the harddrive manufacturers use one definition, and the rest of us another...

The difference, or "tricking", is nothing more than different architectures, and how well they do certain ops... Intel chose one route which gives higher Mhz, but does "less" per Mhz, whereas AMD chose the other route where you get "more" per Mhz...

Thus the market was faced with a choice... Do we continue to rely solely on Mhz when buying a cpu, or do we look at how fast the cpu can do the ops we're interested in?
But AMD chose another route. Instead of teaching the market the differences between the two architectures (which would've been bloody hard to say the least), they choose to create a rating that should show how "fast" their cpu's operates compared to Intel...

Trouble is that this rating isn't set by an "objective" party (there is no such thing as an objective party, but that's another discussion) , but rather something that AMD is doing internally... And so we're stuck with a rating that doesn't tell the consumer how fast the cpu is in Mhz, nor reliably tells him/her how fast it is compared to Intel's cpu's....
(It does give the consumer an idea of how it would equate to Intel's cpu, but it isn't very acurate, and thus not reliable....)

Originally posted by acidosmosis After all, it wasn't AMD that started using these ratings. They were pretty much forced to use them because their processors bear a lower MHz rating than Intel's equivelant's, but that in no way means they perform any worse.

Excuse me?!? AMD didn't start using these ratings first? Then who did?
I can guarranty you that it wasn't Intel...
And neither the RISC's or ALPHA's had such a rating...

You are in some ways correct when you say

their processors bear a lower MHz rating than Intel's equivelant's, but that in no way means they perform any worse.

as in certain ops AMD cpu's are more powerful than Intel's, but not in every op... Just check a review that uses SiSoft Sandra to see what I mean...


Ok, sorry for taking this post so far OT, but I felt a little clarification was needed... Anyways, back to the topic...

Originally posted by olefarte
The idea seems to be that people can buy a 32-bit module, and then add in the 64-bit processor.

Cool! Kinda takes me back to the 486 SX era where you could buy a math-coprocessor...
It would allow you to buy a cheap 64-bit cpu and add a fast 32-bit, since you'd most likely won't need a very powerful 64-bit cpu for a while yet...

Allmost makes me think about switching back to Intel...
(But they're still too expensive....!)

 

[acidosmosis]

Originally posted by MrGaribaldi
Excuse me?!? AMD didn't start using these ratings first? Then who did?
I can guarranty you that it wasn't Intel...
And neither the RISC's or ALPHA's had such a rating...

http://bwrc.eecs.berkeley.edu/CIC/otherpr/prating.html
http://www.pcguide.com/ref/cpu/arch/perfPR-c.html

The Processor Performance Rating (P-rating) specification defines a methodology created by several companies including: AMD, Cyrix, IBM Microelectronics, and SGS-Thomson.

You guys have go to understand AMD's situation. If you jump into the market and go against Intel you really have no choice but to use such a rating. Some of you say well they should show consumers that MHz doesn't mean this or that but that is ignorant. In a way that is what they are trying to do with the p-rating (even though as I said they did NOT create the p-rating, although they did create the new numbers such as 2600).

Why dont you go out and try to teach that to about 5 regular people that know nothing about computers and then 6 months later tell me if they aren't still talking about Intel just because of word of mouth and what TV says. Even though TV has such a large effect on our decisions it wouldn't be easy for AMD to do what they would need to do.

AMD would have to put out some very crafty advertising and prove the facts. Whether you show people facts or not it usually doesnt matter. Take ATI vs Nvidia fans. Now that ATI is ahead and offers more because of DX9 problems with Nvidias card's do you think these Nvidia fans are going to go out and buy an ATI card? No way! And I know you Nvidia fans reading this right now are thinking "no way!". You learn to like something and it gets hard wired into your skull and their is no removing it, at least not easily. That also goes for word of mouth. If you hear it from someone else you automatically think it is good.

When people go out and buy a PC they see this MHz #. If AMD showed that 2.0GHz number compared to Intel's PC running at about 2.8GHz and they are both about the same price (maybe give or take $100-150) which one do you think they are going to buy? INTEL. So you know what if I was AMD I would be doing the exact same thing, using the p-rating. I wouldn't have any choice.

That is one problem that AMD faces and that is the point I was trying to make.

 

[MrGaribaldi]

Originally posted by acidosmosis
http://bwrc.eecs.berkeley.edu/CIC/otherpr/prating.html
http://www.pcguide.com/ref/cpu/arch/perfPR-c.html
[/qoute]

Taken from http://bwrc.eecs.berkeley.edu/CIC/otherpr/prating.htm
Issued jointly by Advanced Micro Devices, Cyrix Corporation, IBM Microelectronics Division, SGS-Thomson

Ok, so AMD did it jointly... But if you read the links, you'll see that the P-rating that AMD is currently using is not following the guidelines set out in the links above, as it is not given equals to a certain cpu, but rather to a hypothetical Intel cpu at that speed...

And even if they were following the guidelines exactly, it still wouldn't give you an "apple to apple" comparison, since

Taken from http://www.pcguide.com/ref/cpu/arch/perfPR-c.html
Warning: The "P rating" is based on system benchmarks that concentrate mostly on integer performance, not floating point. This is important to remember, because many Intel-compatible processors have far inferior floating point units compared to the Intel processors that they claim to be equivalent to with their P rating, and will not perform based on their P rating in floating-point-intensive applications.

Which is partly the point I was trying to make...

<snip>

Originally posted by acidosmosis
That is one problem that AMD faces and that is the point I was trying to make.

That is very well, and I understand where you come from, nor am I opposed to what you're trying to say, though I do find they way you're saying it faulty...

This is the quote that made me write the last post

Originally posted by acidosmosis
Intel is the one doing the tricking. I mean, AMD is compensating for Intel's BS marketing by using a number that will show how well the processor performs as compared to Intel chips.

Here you're saying that Intel is tricking consumers since they (Intel) decided to go with an architecture that gives higher speed in Mhz, but aren't as efficient per Mhz compared to AMD... That is akin to saying it's ATI's fault that nVidia is cheating in benchmarks & hardmuscleing (s?) game dev's because they (ATI) created such a good line-up (R3xx)...

But why don't we lay the matter to rest?

We both agree that AMD cpu's give the most bang-for-buck compared to Intel and that AMD cpu's are able to do more op's (except floating point) per Mhz than Intel, which to me is at the heart of the discussion.

Where we're in disagreement is where you say that AMD was forced to use the PR ratings, whilst I think that was only one of the options available to them, and not the best one at that...

Or we could create another thread to discuss the matter, and not clutter up this thread anymore...


BTW: Just for the record: The last Intel rig I owned/bought was an Intel Celeron 450Mhz... Since then I've had (at least) 3 AMD rigs, and built (at least) 3 AMD rigs for others... So I'm not exactly what you'd call an Intel Zealot....

 

[acidosmosis]

Yea I know what your saying. I didn't word that correctly. Seems that I do that a lot. I apologize for that and you understand my point now so that's good.

 

[consie89]

cpu names? why the hell are you bothered? if its a 5 gig beast whos bothered, and i think adding a number is cool, doing anything else to it would probably make the name cheesy, and we all know how good yanks are at making things sound cheesy.

 

[hunter-killer25]

4000 mhz fsb, god that would have to mean 8 500mhz pumps? yea right or 4 1000mhz pumps? tech doesn't advance that quickly.

 

[agissi]

Originally posted by hunter-killer25
4000 mhz fsb, god that would have to mean 8 500mhz pumps? yea right or 4 1000mhz pumps? tech doesn't advance that quickly.

Wrong.

Yes it does.

It does now.

 

[Didou]

Not really Agissi, it's really hard to have such speeds on electrical circuits. At high speeds, you have lots of noise disturbances, lots of parasite effects & not to mention heat emissions.

Unless they multiplex the signal more then they do now ( by using some 8x pump bus rather then just the Quad-pump bus ), I don't think we'll see those specs for a while now.

 

[agissi]

Originally posted by Didou
Not really Agissi, it's really hard to have such speeds on electrical circuits. At high speeds, you have lots of noise disturbances, lots of parasite effects & not to mention heat emissions.

Unless they multiplex the signal more then they do now ( by using some 8x pump bus rather then just the Quad-pump bus ), I don't think we'll see those specs for a while now.

maybe they well use 8x pump bus? It seems like the industry is growing faster and faster, leaping further and further each time. We went from what.. 1GHz P3, to a 3GHz P4? Why couldnt they leap from 3, to 5GHz? I know theres lots of problems that you run into when you start going this high/fast, but you never know what new technology has been made to go around such obsticals.

 

[Justin]

AMD did not create the P-rating system themselves, but they were the first company to use it in mass-market. They have been using it since fourth generation chips, such as a 5k86 (which was a 486 class chip that included a few pentium extensions).