Intel's Dothan

darko21

Distinguished
Sep 15, 2003
1,098
0
19,280
“””””In a direct comparison between the old Pentium M 1.7 GHz and the new Dothan with 2.0 GHz, the newcomer clearly manages to gain the upper hand. In some of the benchmarks, the mobile CPU produced with 90-nm technology is up to 22% faster. Even if you only consider the difference in clock speed between the two CPUs, Dothan still offers a 5% advantage.””””””

I'm not 100% sure what they are saying a 1.7 giz Dothan is 5% faster then 1.7 giz pentium M???? If that is the case 5% that's decent improvement on a clock for clock comparison.


I did have a problem with this though.

“””””””And again, we applaud Intel's decision to end its marketing strategy of emphasizing gigahertz clock speed specifications as the true measure of processor performance. Nevertheless, it remains to be seen whether the new processor-number style nomenclature will really make it easier for end users to find the right product.””””””

Applaud intel? IMHO intel is changing cause the game is over p4 hit a wall. The MHz myth was intentionally implemented or done by intel. 3 years later this decision based on misleading marketing and ignoring design engineers in favor of marketing to fool or mislead consumers. Why the pat on the back? It's not as if intel changed because of a conscience or something. P4 is finished p3 is what needs to be worked on. Obviously going back to MHz would make intel look bad compared to amd so go with even a more confusing setup based on really (not too much) from a consumers point of view.

A standard set of real benchmarks are what is needed in the cpu industry so consumers are not misled. AMD was pushing for this for the last couple of years but intel will have no part of it. So amd had to drop it.

What I don't understand is why THG is patting Intel on the back.


If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.
 

P4Man

Distinguished
Feb 6, 2004
2,305
0
19,780
>I'm not 100% sure what they are saying a 1.7 giz Dothan is
>5% faster then 1.7 giz pentium M????

Yes, that is their claim. i'll wait for some better reviews before judging, but if this is confirmed, its underwhelming if not unsurprising. Banias already had a large cache, and diminishing returns is what you'd expect from doubling it again. If Dothan is indeed going to be the basis for next generation desktop parts (*), intel better get that clock up significantly, otherwise its desktop future looks bleak. Upping the FSB might help, but considering the huge cache, and the fact the P3 core isnt all that bandwith hungry, I doubt it will result in anywhere as significant speedups as P4A->B->C. Besides, a dual core chip with 2x the fsb/memory bandwith will still only have the same bandwith/core as the current tested Dothan.

(*) anyone else noted the strange form of the die ? You really would'nt want such a rectangular form, unless its already engineered for multicore. Put two of them together, and you get a nice square again (although a rather big one).

>The MHz myth was intentionally implemented or done by intel

No, I disagree here. it has historically become a measure of performance (and a pretty accurate one for a long time as well) even without intel 'intentional efforts'. However, it is obvious intel tried to take advantage of this misconception when they designed the P4. They didnt create the myth, they just abused it :)

>What I don't understand is why THG is patting Intel on the
>back.

Who cares what THG says about it.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
 

Mephistopheles

Distinguished
Feb 10, 2003
2,444
0
19,780
THG's review is a pathetic one. It is not representative of anything, sorry. As for that x86-secret site: Let's refrase that question, shall we? Is this site <i>more reliable</i> than THG? I visit it from time to time, and found no problem with it at all!

<i><font color=red>You never change the existing reality by fighting it. Instead, create a new model that makes the old one obsolete</font color=red> - Buckminster Fuller </i>
 

Mephistopheles

Distinguished
Feb 10, 2003
2,444
0
19,780
Actually, I was reading x86-secret's review, and a 1.7Ghz Dothan sometimes manages a 15+% lead in performance over a 1.7Ghz Banias! How does that leave THG's review? In x86-secret's review, the 2Ghz Dothan even manages to rub shoulders with the heavyweights, like an A64 2.0Ghz and 3+Ghz P4s!

Can you imagine two of these working on a single chip? At 2+Ghz frequencies, with a highly improved memory subsystem? (keep in mind Dothan is only using DDR333 single channel right now)

I'm more inclined to think that THG's review was incomplete. I'm just sorry that there seems to be little information on Dothan right now on the web; I was expecting Xbitlabs and Anandtech to post more thorough reviews. :frown:
 

TheRod

Distinguished
Aug 2, 2002
2,031
0
19,780
The problem is that mobile processor are harder to compare, They often come only in notebooks that can't be tweak/configured to give fair comparison. If all mobile processors were pin to pin compatible with desktop counterpart, it will be easier to review these CPU.

I'm very impressed by the Dothan performance, I think Intel learned from the competition. IBM and AMD turn sooner than Intel into the "non-MHz" path and this paid off. Now that Intel is catching up, AMD must check is back. If Intel turn Dothan to desktop soon, this might "kill" the AMD64 line.

But I doubt Intel will kick Prescott out, they need to make some money out of this "last high-speed" CPU.

--
Lookin' to fill that <font color=blue>GOD</font color=blue> shape hole!
 

darko21

Distinguished
Sep 15, 2003
1,098
0
19,280
Re: No, I disagree here. it has historically become a measure of performance (and a pretty accurate one for a long time as well) even without intel 'intentional efforts'. However, it is obvious intel tried to take advantage of this misconception when they designed the P4. They didnt create the myth, they just abused it :)

Fair enough, I think we are on the same page more or less. Sure the myth existed before the p4 but it was not nearly as pronounced untill the p4. I don't even remember hearing the phrase MHz myth untill the p4 came into existance. Intel took what most people believed all important MHz's and designed around that rather than performance.


If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.
 

P4Man

Distinguished
Feb 6, 2004
2,305
0
19,780
>ctually, I was reading x86-secret's review, and a 1.7Ghz
>Dothan sometimes manages a 15+% lead in performance over a
>1.7Ghz Banias!

Not in any realworld app. Superpi is completely useless as a benchmark (and hardly realworld), and Kribibench can show very weird results.

>n x86-secret's review, the 2Ghz Dothan even manages to rub
>shoulders with the heavyweights, like an A64 2.0Ghz and
>3+Ghz P4s!

Yeah, it performs roughly on par with equivalent clocked A64's. I'm not sure that is so very impressive though (for a future desktop platform), 2 GHz was the the introduction speed of 130nm A64's, 2 GHz for now is the maximum clock for 90nm Dothans (granted, mobile and low power, but still..)

>Can you imagine two of these working on a single chip? At
>2+Ghz frequencies, with a highly improved memory subsystem?
>(keep in mind Dothan is only using DDR333 single channel
>right now)

Can you imagine two A64's cores working on a single chip? 2+Ghz frequencies, with a highly improved memory subsystem?
(keep in mind A64 is only using DDR400 single channel
right now)

;)

Besides as already discussed at length, dual core is probably nice, but definately not the 'end all' of performance. I stand by my point, if intel can't ramp Dothan clockspeeds to match K8's speeds on 90nm, it will be trailing, not leading. And it might be harder to catch K8 on clock as you'd think. Sure the 90nm ones overclock to 2.4, but those speeds are available from AMD as stockspeed on 130nm. Let's see what improvements 90nm bring for AMD before being overjoyed :)

= The views stated herein are my personal views, and not necessarily the views of my wife. =
 

P4Man

Distinguished
Feb 6, 2004
2,305
0
19,780
btw, according to the INQ, dothan will get a 533 MHz bus by the end of this year (no surprise there), and scale to an astonishing 2.13 GHz by Q105
<A HREF="http://www.theinquirer.net/?article=15176" target="_new">http://www.theinquirer.net/?article=15176</A>

= The views stated herein are my personal views, and not necessarily the views of my wife. =
 

P4Man

Distinguished
Feb 6, 2004
2,305
0
19,780
> As for that x86-secret site: Let's refrase that question,
>shall we? Is this site more reliable than THG?

You mean that site that refuted the INQ's claims Tejas was canned ?

:)

Actually, I think x86-secret and THG would run neck and neck when it comes to cluelessnes, poor benchmark selections and irrelevant (not to say downright idiotic at times) "analysis".

= The views stated herein are my personal views, and not necessarily the views of my wife. =
 

trooper11

Distinguished
Feb 4, 2004
758
0
18,980
yeah i have to agree with you there. the benchamrks were very limited, throwing up just a few benches. the only one i found interesting at all was the divx one, which was nice. but i want see the whoel thing, including gaming benchmarks, but i guess we wont be able to ever compare till dothans come to the desktop , since the notebooks are limited in power.
 

Mephistopheles

Distinguished
Feb 10, 2003
2,444
0
19,780
It wasn't just them who refuted the INQ's claims that Tejas was canned. Aceshardware did that as well.

In any case, take it easy, P4Man (poor name... heh)... we all need more information on that processor in order to be truly informed about its potential.

<i><font color=red>You never change the existing reality by fighting it. Instead, create a new model that makes the old one obsolete</font color=red> - Buckminster Fuller </i>
 

P4Man

Distinguished
Feb 6, 2004
2,305
0
19,780
>It wasn't just them who refuted the INQ's claims that Tejas
>was canned. Aceshardware did that as well.

Nope, they didnt afak. Aces isn't much of a news/rumour reporting site like the INQ, Xbit or x86-secret anyway; they don't have any inside sources afaik -besides Groo (Charlie) from the INQ posting regulary in the forum, but usually after the INQ article is up :)

>we all need more information on that processor in order to
>be truly informed about its potential.

Well, potential can only be guestimated no matter how much info we have :) But I think we have enough to make educated guesses. Dothan seems to perform roughly on par with A64 per clock, question remains how high they can clock it, and how much higher fsb will help. I guess I'm more pessismistic than most on both accounts (and not to tooth my own horn, but I'm rarely very wrong in my guestimates).

Dual core will come from both AMD and intel at roughly the same time, I even expect at least opterons to be out before any dothan based dual core, and they will pretty much perform like a 2 way SMP system, hence, not really that impressive compared to single cpu systems on most current (desktop) software. If, when or how much future games/desktop apps will benefit remains an open question as well, but also there I'm more pessimistic as most: rendering, video/audio encoding and photoshop filter type workloads will gain substantially (they mostly already do from SMP), but games, compiling and many other software are not likely to become all that SMP friendly within the next years -if ever. Parallelizing software just ain't that easy, and the benefits don't always outweigh the disadvantages (overhead).

= The views stated herein are my personal views, and not necessarily the views of my wife. =
 

Crashman

Polypheme
Former Staff
Yes, Intel coppied AMD's idea, but using different numbers from the Opterons and FX's. Terrible when Intel does it, isn't it?

<font color=blue>If you spilled a box of toothpicks on the floor, could I tell you how many there are?<font color=blue>
<font color=red>No, but what does that matter, I got a free Motherboard!</font color=red>
 

darko21

Distinguished
Sep 15, 2003
1,098
0
19,280
I'm not so sure I see your point. In the server market multiple cpus can be used and naming games are no where near as effective…I-T's should be some what educated. I think it's fair to assume when amd was pushing for a benchmark standard in the industry amd would have been happy to accommodate any server chip issues intel felt important. Intel wanted no part of this and the average consumer would be much better off with it.

I can't say I blame intel when you look at what a standard set of performance measuring benchmarks would expose. Price performance and intels reputation would be at risk. Intel cannot have that as only the consumer would benefit from that.

AMD did not push for a numbering scheme they wanted honest benchmarks to dictate a cpu's value. Intel had every chance to accept it but intel and companies like dell would have no part of it. and amd was forced to drop the proposal. (they fought for at least two years) to no avail.

Crash do you really believe intel is copying amd's high-end numbering system? If so why not use amd numbers to rightly compare. No I personally don't think you believe that. I think you should be intelligent enough to understand the p4 design has collapsed and intel will be relying on the p3 design to move forward. Since this would put intel in a marketing disadvantage have fooled consumers for the last 3 years with MHz. So bring on a set of meaning less numbers for the consumer cause they do not want the average consumer to understand the truth.

At least that’s my take on it.


If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.
 

Crashman

Polypheme
Former Staff
It's the same story either way you look at it. Opteron 1.6GHz was a fast processor, calling it 1.6GHz would have only served to reduce sales because of the name implying it was "slower" than a Xeon. So they introduced a new number that wasn't based on anything, except for comparing the processor to other speeds of the same processor.

Intel's doing the same thing for the same reason. Same logic. AMD did it first, therefor anyone can claim Intel might have gotten the idea from AMD.

It works, or doesn't, the same way. FX51/53/whatever, you assume the 53 is faster than the 51, but what does that compare to with other processors? Nothing, it's not supposed to. It simply implies the 53 is faster than the 51, which it is. No dishonest or anything!

So AMD's FX and Opteron numbers can't be compared to other processors, and now Intel's setting up the same type of scheme. If Intel had used AMD's numbers, they'd have a couple problems: 1.) Everyone would know for SURE they coppied AMD completely, 2.) It would be up to them to maintain honesty, just as it was up to AMD to maintain honesty on the XP rating system.

Nope, AMD's model numbers don't compare to other processor families, and neither will Intel's. Same deal, different numbers.

<font color=blue>If you spilled a box of toothpicks on the floor, could I tell you how many there are?<font color=blue>
<font color=red>No, but what does that matter, I got a free Motherboard!</font color=red>
 

darko21

Distinguished
Sep 15, 2003
1,098
0
19,280
Once again I don't see your point.

Re: It's the same story either way you look at it.

Only cause intel refused to accept a benchmark standard.

Re: Opteron 1.6GHz was a fast processor, calling it 1.6GHz would have only served to reduce sales because of the name implying it was "slower" than a Xeon.

As I said I'm reasonably sure amd would have been happy to use opteron in a real benchmarking scheme. and your not fooling very many in this category cause for the most part they are I-T's who know what questions to ask.

Re: So they introduced a new number that wasn't based on anything, except for comparing the processor to other speeds of the same processor.

Yes a pity really, but why not as multiple cpu's can be used and intel refused a benchmarking strategy. Which is what the industry needs.

Re: Intel's doing the same thing for the same reason.

No here is where you are out to lunch. AMD is doing it in the high end cause intel refused to accept a benchmarking setup. Intel is doing it cause the p4 failed and the MHz myth is over for them. Surley you can see that. Do you really expect people do believe that if prescott was @ 4 giz 90 nm right now intel would be adopting a number scheme? Nope dothan would be moblile with no talk of a desktop version and prescott would be everything and a number scheme out of the question. MHz rules.

Re: Same logic.

No here is where you are out to lunch. AMD is doing it in the high end cause intel refused to accept a benchmarking setup. Intel is doing it cause the p4 failed and the MHz myth is over for them. Surely you can see that. Do you really expect people do believe that if prescott was @ 4 giz 90 nm right now intel would be adopting a number scheme? Nope dothan would be moblile with no talk of a desktop version and prescott would be everything and a number scheme out of the question. MHz rules.

Re: AMD did it first, therefore anyone can claim Intel might have gotten the idea from AMD.

Lets wake up and face reality head on shall we. Intel refused a standard set of benchmarks to dictate a cpu's overall performance level, correct. If intel won't do what's best for the consumer why should AMD use MHz or a rating comparison to the p4 when so many other intel chips are around. MHz means zip and you can't compare when when so many different chips performing differently at different clocks.

Re: It works, or doesn't

It doesn’t work for the consumer but who's fault is that?


Re: the same way. FX51/53/whatever

Yup, but who's fault is that?


Re: you assume the 53 is faster than the 51, but what does that compare to with other processors? Nothing, it's not supposed to.

Yup, but who's fault is that? Do you believe amd wanted a rating formula based on benchmarks similar to the xp-rating formula but the fx cpu's must have a separate numbering system? Or do you think fx should just mention MHz which means zip. No amd had the right idea too bad intel refused to do what was best For the consumer.

Re: It simply implies the 53 is faster than the 51, which it is. No dishonest or anything!

Yes lets hope intels number system is at least that honest. Although I'd bet my left nut intel gives a slower performing CPU a greater number than a faster fx cpu. Or would you disagree with that as well. Intel does not wish to be compared to AMD on any performance level. Why do you think that is???? Toyota's compare them selves to Nissans GM's to Honda's etc etc. Competition is good for the consumer so they understand value.

Re: So AMD's FX and Opteron numbers can't be compared to other processors

No they can't but whose fault is that?


Re: and now Intel's setting up the same type of scheme.

No here is where you are out to lunch. AMD is doing it in the high end cause intel refused to accept a benchmarking setup. Intel is doing it cause the p4 failed and the MHz myth is over for them. Surely you can see that. Do you really expect people do believe that if prescott was @ 4 giz 90 nm right now intel would be adopting a number scheme? Nope dothan would be moblile with no talk of a desktop version and prescott would be everything and a number scheme out of the question. MHz rules.

Re: If Intel had used AMD's numbers, they'd have a couple problems: 1.) Everyone would know for SURE they coppied AMD completely, 2.) It would be up to them to maintain honesty,


Intel did not have to use AMD numbers... AMD was try for at least the last 2 years to set up an industry wide set of benchmarks to prove a CPU's overall value based on standard benchmarks. Why do you think intel refused this?


Re: just as it was up to AMD to maintain honesty on the XP rating system.

And that AMD did... Don't make me go though this again cause you should know by now I will.


Re: Nope, AMD's model numbers don't compare to other processor families, and neither will Intel's. Same deal, different numbers.

They don't but give AMD credit where credit is due. The xp rating formula was the closest we ever got while overly generous to begin with more accurate in the middle and failing in the end it was a valid attempt to compare one on one. Crash just think of how much more accurate te xp rating formula would have been if intel used the same formula.

Intel had every opportunity to adopt a true benchmarking formula like AMD pushed so hard for, but Intel will have no part of it so crash why do you think that is?


If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.
 

sonoran

Distinguished
Jun 21, 2002
315
0
18,790
RE: "pushing for a benchmark standard in the industry"

darko, what do you think <A HREF="http://www.tpc.org/information/who/whoweare.asp" target="_new">this</A> is? If people don't agree on those benchmarks (even though <b>all</b> the major players are members), why are they going to agree on another set?
 

darko21

Distinguished
Sep 15, 2003
1,098
0
19,280
What is your point?

How many average consumers view that stuff?

Use those benchmarks whatever but let the consumer know how it preforms so they can compare value. Just like cars HP or MPG, the information could be on some obscure web site but consumer needs to know when buying. Not blaming you but is it really that hard to understand?

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart. <P ID="edit"><FONT SIZE=-1><EM>Edited by darko21 on 05/10/04 10:18 PM.</EM></FONT></P>
 

Crashman

Polypheme
Former Staff
But AMD didn't adopt a benchmarking program to come up with model numbers, not because Intel wouldn't agree, but because they didn't. They didn't need Intel's permission, agreement, or aknowledgement, they could have done it on their own. They adopted a new naming system because their "slower" processors were providing better performance. Intel did the same thing. Intel's biggest problem right now is that the PIII based CPU's have always provided better performance per clock cycle than their P4 based processors, and that the Pentium-M is an advanced version of the PIII. It gets kind of hard to sell P-M 1.8's when people don't know they're faster than P4-M 2.6's. But then again, it would have been kind of hard for AMD to sell Opterons at 1.6GHz and convince the uninformed IT guys those were faster than the Xeons, reason being that IT guys are lazy in general and won't read unless they have to. I know IT guys, they'd look at the MHz and assume Intel had the lead, so AMD came out with a number that couldn't be misinterpreted as a MHz reference, forcing the IT guys to look deeper.

Yep, AMD couldn't make higher MHz CPU's and neither could Intel. Reason 2 is they both know they can produce faster performing processors at slower clock rates. So both of them developed numbering schemes that were unrelated to MHz. AMD lead the market, Intel followed. Their reasons were the same.

You can dance around that all you want, simply because Intel was promoting MHz long after AMD gave up on it, but the fact remains Intel had to follow AMD's idea because Intel found out LATER than AMD that they couldn't ramp up clock speed any higher, and their clock speed promos were killing P-M sales.

They did the same thing for the same reasons. AMD could have come up with a standard list of benchmarks and a formula for providing a performance number based on some kind of weighted average, but they didn't. Can't blame Intel, that was AMD's choice. And now Intel is doing the same thing, and for the same reasons AMD has done it.

<font color=blue>If you spilled a box of toothpicks on the floor, could I tell you how many there are?<font color=blue>
<font color=red>No, but what does that matter, I got a free Motherboard!</font color=red>
 

Crashman

Polypheme
Former Staff
Although I'd bet my left nut intel gives a slower performing CPU a greater number than a faster fx cpu.
Why would they? Why wouldn't they? The numbers have nothing to do with each other, just as FX numbers have nothing to do with Opteron numbers. It's going to be left up to consumers now to read reviews and figure this stuff out for themselves. But since customers are lazy, it will likely result in Intel keeping their loyalist, AMD keeping their's, and people in the middle buying whatever's cheapest.

<font color=blue>If you spilled a box of toothpicks on the floor, could I tell you how many there are?<font color=blue>
<font color=red>No, but what does that matter, I got a free Motherboard!</font color=red>
 

juin

Distinguished
May 19, 2001
3,323
0
20,780
Nope, they didnt afak. Aces isn't much of a news/rumour reporting site like the INQ, Xbit or x86-secret anyway; they don't have any inside sources afaik -besides Groo (Charlie) from the INQ posting regulary in the forum, but usually after the INQ article is up


Yes i give the link and it seen they are right you have write just few hour to fast

i need to change useur name.
 

darko21

Distinguished
Sep 15, 2003
1,098
0
19,280
Re: Why would they?

Cause its misleading and its about ripping off the consumer.

Re: Why wouldn't they?

They will.... See reasons above.

Re: The numbers have nothing to do with each other,

Nope but they could have if intel had agreed to a benchmarking strategy.


Re: just as FX numbers have nothing to do with Opteron numbers.

No they don't.. I thought I had explained this already did you even read the reply I made to you?

Re: It's going to be left up to consumers now to read reviews and figure this stuff out for themselves.

Yup but most consumers can't be bothered to look it up and will not realize they were ripped off.

Re: But since customers are lazy, it will likely result in Intel keeping their loyalist,

You mean like kanavite? What about ripping off the uninformed like my dad or you mom who really don't think like kanavite or care amd intel, but just want the best value for there hard earned dollar?

Re: AMD keeping their's,

Loyalists are retards if I thought I'd get more performance from intel for my dollar I'm there. End of story. Granted if tied I'd through my support with amd but only cause the industry needs competition amd offers just that but do to misleading marketing they are struggling.

Re: and people in the middle buying whatever's cheapest.

Like all the uniformed such as our parents buying dells with a 2.7 giz celeron. That's unacceptable in my book.


If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.
 

Crashman

Polypheme
Former Staff
To begin with, Intel's model numbers are 3 digits long, AMD's FX series is 2 digits long, it's apples to oranges, there are no UNITS. If I told you my monitor was 40 wide, you might say "Wow, a 40" wide monitor?", and I'd say No, 40cm. Without units you have no system for basing anything, 3 digits, 2 digits, there's no room for Intel to be "dishonest" there if the numbers don't mean anything.

Your argument about Intel using higher numbers is like saying "Hey, it's unfair to call your bicycle a Mongoose, because mine's a Diamond Back, I know Mongoose eat snakes, but my bikes better than yours". The numbers simply don't relate to anything, they're names. If a Jaguar 120 went 120MPH, would a Chrysler 300 go 300? LOL.

It's the same story, Intel's now doing what AMD's been doing, for the same reasons.

<font color=blue>If you spilled a box of toothpicks on the floor, could I tell you how many there are?<font color=blue>
<font color=red>No, but what does that matter, I got a free Motherboard!</font color=red>
 

darko21

Distinguished
Sep 15, 2003
1,098
0
19,280
Re: But AMD didn't adopt a benchmarking program to come up with model numbers, not because Intel wouldn't agree, but because they didn't.

Could you clarify just what are you saying here.

Re: They didn't need Intel's permission, agreement, or aknowledgement, they could have done it on their own.

I am asumming they means amd? If that is the case don’t be ridiculous. It’s as if you are trying to confuse the topic or something. So lets break down what you just said.

<font color=red> “But AMD didn't adopt a benchmarking program to come up with model numbers” </font color=red> They certainly did with the xp-rating formula and they have certainly been trying to for the last few years. Did I not already explain this fully to you why are you making me repeat it differently. Why do I make these posts and you donb’t even read it? <font color=red> “not because Intel wouldn't agree, but because they didn't. “ </font color=red> So you are on the board of directors now? You know Intel wanted an honest rating formula but did not care for the one amd proposed. I’d bet my right nut this time that amd would have accepted any independent body to decide on the benchies used. <font color=red> They didn't need Intel's permission, agreement, or aknowledgement, they could have done it on their own. </font color=red> Done what on there own? You mean like the xp-rating fonmula? Did you not read my previous posts? If Intel won’t play what is the point? To many different cpu’s speeds ipc this has been fully explained already. The point is for Intel to confuse mislead.



Re: They adopted a new naming system because their "slower" processors were providing better performance.

Who did AMD? It’s not slower its faster just runs at less MHz. and you are WRONG amd went to a naming scheme cause Intel refused to adopt a proper benchmarking scheme. I thought I had explained this already why do you make me repeat it?


Re: Intel did the same thing. Intel's biggest problem right now is that the PIII based CPU's have always provided better performance per clock cycle than their P4 based processors, and that the Pentium-M is an advanced version of the PIII.

Funny how you twist things around. A couple of months ago Intel had the advanced new generation p4 and amd’s k8 was was 1990’s k7 technology. So funny how you phrase it now. But of course you would never see it that way.

Re: It gets kind of hard to sell P-M 1.8's when people don't know they're faster than P4-M 2.6's.

Yes it would be. And its intels own fault for going there to begin with. I think the bigger problem is how would you compare an dothlon @2.2 giz to an fx58 @ 2.8 giz?


Re: But then again, it would have been kind of hard for AMD to sell Opterons at 1.6GHz and convince the uninformed IT guys those were faster than the Xeons

Leave that to the Sun and HP sales reps to explain why it’s superior. They might even buy you a cheeseburger while they explain it.


Re: reason being that IT guys are lazy in general and won't read unless they have to. I know IT guys, they'd look at the MHz and assume Intel had the lead,

You mean like my grand parents? They gotta be better than that.

Re: so AMD came out with a number that couldn't be misinterpreted as a MHz reference, forcing the IT guys to look deeper.

AMD wanted a benchmark rating for the industry Intel said NO THANKYOU. Did you read any of my posts to you?



Re: Yep, AMD couldn't make higher MHz CPU's and neither could Intel.

Please don’t speak for me. For all we know AMD could have made a 5 giz cpu but preformed like a 500MHZ k6. Intel hit a wall with the high MHz low IPC misleading design. And I say let them eat cake.


Re: Reason 2 is they both know they can produce faster performing processors at slower clock rates.

Wrong again only Intel can do that. Because only Intel implemented a misleading design.


Re: So both of them developed numbering schemes that were unrelated to MHz.

And this was the wrong approach. Only implemented cause Intel refused to be compared with realistic benchmarks for a numbering system.


Re: AMD lead the market, Intel followed. Their reasons were the same.

NO how many times do I have to say this????????????
AMD tried for the last few years to implement an industry wide set of benchmarck to rate performance over all. Intel refused. WHY DO YOU THINK THAT IS CRASH
???????? It was only after Intel refused that that they let the whole thing go. What’s the alternative MHz trying to figure out how to rate to a Prescott and p4c xenon and dothon biannas whatever.


Re: You can dance around that all you want

Really I am the one dancing here.


Re: simply because Intel was promoting MHz long after AMD gave up on it

Intel went to high MHz low IPC to mislead not to do anyone any favors. AMD did not give up… they were trying to show the uninform true value as intels high MHz low IPC was only meant to mislead.


Re: but the fact remains Intel had to follow AMD's idea because Intel found out LATER than AMD that they couldn't ramp up clock speed any higher, and their clock speed promos were killing P-M sales.

NO how many times do I have to say this????????????
AMD tried for the last few years to implement an industry wide set of benchmark to rate performance over all. Intel refused. WHY DO YOU THINK THAT IS CRASH
???????? It was only after Intel refused that that they let the whole thing go. Whats the alternative MHz trying to figure out how to rate to a Prescott and p4c xenon and dothon biannas whatever.


Re: They did the same thing for the same reasons.

WRONG amd went to a naming scheme cause Intel refused to adopt a proper benchmarking scheme. I thought I had explained this already why do you make me repeat it?

Re: AMD could have come up with a standard list of benchmarks and a formula for providing a performance number based on some kind of weighted average, but they didn't.

AMD did with the xp and were trying for the last 2 years at least for any future cpu’s but big wigs like dell Intel won’t have it. . WHY DO YOU THINK THAT IS CRASH


Can't blame Intel, that was AMD's choice.

Yes you most certainly can. Just what couor is the sky in your world anyway? Intel could have adopted a rating formula based on benchies. As a matter of fact if I had a third nut I bet amd would have let Intel pick all the benchies so long as Intel agreed to adopt it for 10 years or so. But Intel will have no part of it.


Re: And now Intel is doing the same thing, and for the same reasons AMD has done it.

NO how many times do I have to say this????????????
AMD tried for the last few years to implement an industry wide set of benchmark to rate performance over all. Intel refused. WHY DO YOU THINK THAT IS CRASH
???????? It was only after Intel refused that that they let the whole thing go. Whats the alternative MHz trying to figure out how to rate to a Prescott and p4c xenon and dothon biannas whatever.




If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.