Prescott or P4?

[Ticktanium2038]

Not looking to generate a War just looking for reasonable answers.

Given recent announcements. Prescott's highest clockspeed with be at 3.4gig in the 1st quarter of next year. It will not reach 3.6 until the second half of next year.

According to benchmarks floating around the web. A prescott at 3.4 will perform the same as a P4 at 3.2. The only thing prescott will add is SSE3 (Which Nothing is Programmed for except benchmark utilities)

Why would anyone buy a prescott when you could buy a P4 2.4C and overclock to 3.2 or higher? It will be cheaper and produce a lot less heat giving you a quieter PC.

Its also said that overclocking a Prescott to 3.6 will cause it to burn up and die. So I dont expect it to be a good overclocker. Dont ask how I know that this is not what this is about.

Please also refrain from Answering with Buy an Opteron. I want people who support Intel to give reasonable answers to why they would buy a prescott or if they will wait until the prescott gets better.

 

[Acanthus]

The prescott on the .09 process SHOULD run cooler than its .13 northwood counterpart. It runs on a lower vcore. I would imagine that overclocking would not hurt the CPU without increasing the vcore. But its almost certain that they will be poor overclockers upon release (most cpus are).

 

[Ticktanium2038]

Its my understanding this Prescott will be running at 103 Watts. Unforuntatly it wont be much lower. So this Prescott might be a waste.

They are revamping/Rearranging the core so that they can get the CPU above 3.6ghz and solve some of the problems they are having with heat and leakage. I would expect up to 3.8 out of the revamped core and the 3.6 POP GOES THE WEASEL to be under control then.

 

[bgeh]

maybe Intel won't include strained silicon in the first batches of the Prescott, that should account for the heat and low clockspeeds. once it's implemented, it should be able to go up

 

[bgeh]

sorry, dp

 

[Ticktanium2038]

Does anyone know when Intel plans on using their strained silicon? This would indeed help the leakage problem. Does anyone know or have links to where Intel stands on their Strained silicon process? Or is this not until Tejas.

I have to say I have nothing about thier Strained silicon only that I recall reading about it.

 

[DAPUNISHER]

As Wingznut stated, the Opteron looked less than spectacular when the ES models were being evaluated but performed much better as the shipped version, the same will most likely hold true for Prescott. If the roadmap for pricing some members linked awhile back holds up then by Feb. the Prescott 2.8 will be sub 180$, should it be capable of 3.6+ghz on air, and the HT has the niffty little tweaks that are rumored, plus 3D rendering and Video encoding software be available with SSE3 support, and a new chipset maybe? Then it's going to be a great price/performer in the areas where Intel is already very strong. Certainly there are some rather large "if's" in there but historically Intel can fall in a pile of dog poop and come out smelling like a rose E.G. P4 improvements over the course of it's life cycle.

 

[Ticktanium2038]

The difference then was the Opteron was only running at 800Mhz and not at full speed.

Weve seen the Prescott at 2.8 and it doesnt perform as well as the P4 at 2.8. Except when SSE3 instructions are added of course. I am sure that those are legit benchmarks of the performance of the chip at that speed.

This was the same for the P3 when the P4 was launched the P4 was a slower processor clock per clock but it was designed in a way that it could ramp up clock speed much higher than the p3 eventually making it a better CPU although the P3 was better designed. Probably Intels best CPU design ever in my eyes.

I see prescott as the same but what I dont see is Intel being able to ramp up speeds as easily as the P4 days. For steady streaming applications like video, audio encoding I see it doing well but because of the deeper pipelines I see it as a problem when doing other things like games and databases.

For the money factor I still see the Intel P4 2.4C 800FSB overclocked to 3.2 or higher as being the best intel bang for the buck. Even better than Prescott.

So I am not sure where prescott is going unless Intel really believes they can ramp up clock speeds well beyond 4gig in the near future with Xeon type chips doing the same? But I dont see that any time soon.

Might it be better for Intel to current port the P4 to the 9 micron instead of trying to work out the problems with Prescott? Maybe SSE3 is part of the hot spot problem? Maybe the deeper pipelines are contributing to the current leakage? My guess is only the engineers know?

 

[sao123]

Well my first intuition is this... wait for the socket 775LGA prescotts. (BTW...this is when strained silicon is supposed to be introduced) They are supposed to debut Q3 04. This guarantees you arent buying a end of the roadmap mobo & cpu. IE...in the near future AGP will be replaced with PCIX and DDR II, also SATA is supposed to get a bump to 300MB by Q1 05 at the latest.

As to answer your question, The only reason why a 3.4prescott would only perform as a 3.2 northwood is that the software optimazation isnt there yet. It will be soon. Secondly, by doubling the L2 cache (adding 50million more transistors) thats why it runs so hot. I dont think it unreasonable to run a 3.4 @ 3.6 without too much trouble, though you might want water cool, because the same yeild will be producing both, but the ones certified for 3.6 is much smaller than most present top of the line yields. But my first intuition is if your gonna buy a prescott to over clock... why not get the 2.8? Its only gonna be around $220.

Another reason why to get the prescott over the northwood? Improved hyperthreading, deeper pipe and better branch prediction, SSE3.


Now let me offer some of my own thoughts...worrying about .2Ghz worth of CPU speed aint worth it....there aint gonna be much overall noticable difference between even a 2.8 prescott and 3.4 prescott. why worry about the speed grade of your cpu? The cpu is the fastest piece of hardware in your system, and probably the most under-utilized. Your pc is going to only go as fast as your slowest hardware. Put crappy video, not enough RAM, or 5400 rpm parallel ata100 drives in your system and watch it crawl.

If you want real performacne...upgrade the other critical areas of a PC.
1)Not enough RAM will slow your system to a halt, spring for 512GB or 1GB even of DDR 400.
2)Hard Drives tend to slow down PC's more than most anything else (except #1) because of the nature of the windows swap file. Go SATA with your extra money. Heck, go with 10000 rpm SATAs from WD. Go SCSI. Use raid 0.
3) The consider a top end video card, with 128 or 256MB memory.

sorry bout going off topic.

 

[beyoku]

Originally posted by: Acanthus
The prescott on the .09 process SHOULD run cooler than its .13 northwood counterpart. It runs on a lower vcore. I would imagine that overclocking would not hurt the CPU without increasing the vcore. But its almost certain that they will be poor overclockers upon release (most cpus are).

"SHOULD" is why they are having trouble in the first place LOL

 

[beyoku]

This was the same for the P3 when the P4 was launched the P4 was a slower processor clock per clock but it was designed in a way that it could ramp up clock speed much higher than the p3 eventually making it a better CPU although the P3 was better designed. Probably Intels best CPU design ever in my eyes.

i dont really think that is true - the only think that actually made the p4 better was frequency - that is it. SO if clock frequency makes a "better" CPU then you may have a case. But i dont think it was actually "better" it just had speed.

 

[jiffylube1024]

Originally posted by: Ticktanium2038
Not looking to generate a War just looking for reasonable answers.

Given recent announcements. Prescott's highest clockspeed with be at 3.4gig in the 1st quarter of next year. It will not reach 3.6 until the second half of next year.

According to benchmarks floating around the web. A prescott at 3.4 will perform the same as a P4 at 3.2. The only thing prescott will add is SSE3 (Which Nothing is Programmed for except benchmark utilities)

Why would anyone buy a prescott when you could buy a P4 2.4C and overclock to 3.2 or higher? It will be cheaper and produce a lot less heat giving you a quieter PC.

Its also said that overclocking a Prescott to 3.6 will cause it to burn up and die. So I dont expect it to be a good overclocker. Dont ask how I know that this is not what this is about.

Please also refrain from Answering with Buy an Opteron. I want people who support Intel to give reasonable answers to why they would buy a prescott or if they will wait until the prescott gets better.

If this is your logic (don't buy ahead for SSE3) then why would anyone buy a 64-bit AMD CPU either?

And everything about Prescott is just speculation right now. Intel won't release the new stepping until it's ready, and until it's faster than what's out there already. Especially since they won't give it a name like the "Pentium V" now (it's going to be under the P4 name).

It's obvious by now that Prescott has been, up to this point, a disaster. If they were to release it in it's current form it would be a train wreck. The writing is on the wall with the launch delays, the demoing of the chip a a paltry 2.8Ghz (well below .13um P4's), etc. And the unfair marketing tactics that is the P4EE signals they're desparate to get something out the door to put a stop to AMD Athlon64 chips.

However, why would you even bring up a thread topic like "Prescott or P4" in the CPU forum when it has been confirmed that Prescott will be released in January/February at the earliest? We know Prescott is late to the dinner party, and we know Intel has been having problems. Moreover you have brought up this same issue before. Why rub everyone's nose in it?

In it's current form, Prescott sucks. When it's launched, I wouldn't be surprised if it scales well and takes the P4 to new heights. It will probably take a few months into 2004, but I'm sure Intel will get Prescott or it's sucessor out the door and in good shape.

 

[DAPUNISHER]

Now let me offer some of my own thoughts...worrying about .2Ghz worth of CPU speed aint worth it....there aint gonna be much overall noticable difference between even a 2.8 prescott and 3.4 prescott. why worry about the speed grade of your cpu? The cpu is the fastest piece of hardware in your system, and probably the most under-utilized.

That's only accurate to a certain degree, many of us run Distributed Computing projects that load 100% of the areas of the Processor they utilize, and with projects that allow multiple instances or are MP aware they use even more. For myself, I haven't seen any CPU yet that can crunch SETI or F@H nearly fast enough for my liking

This was the same for the P3 when the P4 was launched the P4 was a slower processor clock per clock but it was designed in a way that it could ramp up clock speed much higher than the p3 eventually making it a better CPU although the P3 was better designed. Probably Intels best CPU design ever in my eyes.

The P4 has evolved significantly since the first model arrived, pin count, substantial FSB and clock speed increases, cache size, .18 to. 13, and little fixes like the C1 stepping had that fixed the Data Prefetch bug increasing performance roughly 2%. And no way IMO is the P3 a better design than the P-M though it is evidently partly based on the P3.

 

[Wingznut]

Originally posted by: Ticktanium2038
Not looking to generate a War just looking for reasonable answers.

Yeah, right. That would be SO out of your character.

If that was true, then you'd do some research. You'd know that Prescott uses Strained Silicon, and you'd understand WHAT it is. You'd understand WHY you cannot project performance based on a pre-production stepping. You'd understand that YOU cannot predict the future, yet you make all these predictions as if you actually know. Some examples of your throwing out statements as if they were fact...

• It (Prescott) will not reach 3.6 until the second half of next year.
• Its also said that overclocking a Prescott to 3.6 will cause it to burn up and die.
• A prescott at 3.4 will perform the same as a P4 at 3.2.
• The only thing prescott will add is SSE3...
• This (strained silicon wafers) would indeed help the leakage problem.
• Maybe SSE3 is part of the hot spot problem?
• Maybe the deeper pipelines are contributing to the current leakage?
• Might it be better for Intel to current port the P4 to the 9 micron instead of trying to work out the problems with Prescott?

I've asked this before, and I'll ask it again.... What oriface do you pull these things out of???

 

[User1001]

Originally posted by: Wingznut

Originally posted by: Ticktanium2038
Not looking to generate a War just looking for reasonable answers.

Yeah, right. That would be SO out of your character.

If that was true, then you'd do some research. You'd know that Prescott uses Strained Silicon, and you'd understand WHAT it is. You'd understand WHY you cannot project performance based on a pre-production stepping. You'd understand that YOU cannot predict the future, yet you make all these predictions as if you actually know. Some examples of your throwing out statements as if they were fact...

• It (Prescott) will not reach 3.6 until the second half of next year.
• Its also said that overclocking a Prescott to 3.6 will cause it to burn up and die.
• A prescott at 3.4 will perform the same as a P4 at 3.2.
• The only thing prescott will add is SSE3...
• This (strained silicon wafers) would indeed help the leakage problem.
• Maybe SSE3 is part of the hot spot problem?
• Maybe the deeper pipelines are contributing to the current leakage?
• Might it be better for Intel to current port the P4 to the 9 micron instead of trying to work out the problems with Prescott?

I've asked this before, and I'll ask it again.... What oriface do you pull these things out of???

Wut he said

 

[DivideBYZero]

Originally posted by: Wingznut

Originally posted by: Ticktanium2038
Not looking to generate a War just looking for reasonable answers.

Yeah, right. That would be SO out of your character.

If that was true, then you'd do some research. You'd know that Prescott uses Strained Silicon, and you'd understand WHAT it is. You'd understand WHY you cannot project performance based on a pre-production stepping. You'd understand that YOU cannot predict the future, yet you make all these predictions as if you actually know. Some examples of your throwing out statements as if they were fact...

• It (Prescott) will not reach 3.6 until the second half of next year.
• Its also said that overclocking a Prescott to 3.6 will cause it to burn up and die.
• A prescott at 3.4 will perform the same as a P4 at 3.2.
• The only thing prescott will add is SSE3...
• This (strained silicon wafers) would indeed help the leakage problem.
• Maybe SSE3 is part of the hot spot problem?
• Maybe the deeper pipelines are contributing to the current leakage?
• Might it be better for Intel to current port the P4 to the 9 micron instead of trying to work out the problems with Prescott?

I've asked this before, and I'll ask it again.... What oriface do you pull these things out of???

Owned.

 

[Ticktanium2038]

I think I have made the comment enough times on buying a P4 2.4C enough that I cant be classified as Anti-Intel.

I dont see how prescott helps the Intel Mix ramp up their CPU line.

Generally a next gen processor is a performance increase. I see prescott as nothing more that a slightly more expensive P4 that isnt an improvement over the existing P4 3.2.

I think sao123 had the best comment:
------
Well my first intuition is this... wait for the socket 775LGA prescotts. (BTW...this is when strained silicon is supposed to be introduced) They are supposed to debut Q3 04. This guarantees you arent buying a end of the roadmap mobo & cpu. IE...in the near future AGP will be replaced with PCIX and DDR II, also SATA is supposed to get a bump to 300MB by Q1 05 at the latest.
------

I think Intel releasing Prescott at existing speeds 2.8-3.4 will only make them look stupid when the benchmarks reveal the processor is slower than their existing line at higher speeds. Maybe waiting until the 775LGA where the core will be mature enough to qualify as a superior chip to thier own existing P4 line.

Why does someone want to spend more for something that perform less?

But I guess you couldnt keep your mouth shut wignutz and offer a good investor or technical perspective of the situation. Try answering the question next time instead of personal bashing because I called you out before. Still waiting for your 100+ watt cpu listing you claimed some time ago? Your personal bashing is really lame get a life.

 

[Ticktanium2038]

It (Prescott) will not reach 3.6 until the second half of next year.
**TRY READING INTELS ROADMAP GENIUS ITS ALL THERE FOR YOU**

Its also said that overclocking a Prescott to 3.6 will cause it to burn up and die.
**YUP SOURCES CONFIRMED IT WILL BURN UP**

A prescott at 3.4 will perform the same as a P4 at 3.2.
**TRY A WEB SEARCH PRESCOTT IS LOWER THAN P4 ITS OUT THERE**

The only thing prescott will add is SSE3...
**OK SO IT WILL ADD DEEPER PIPELINES AND BETTER BRANCH PREDICTION**

This (strained silicon wafers) would indeed help the leakage problem.
**NO KIDDING I DIDNT SAY IT WOULDNT I ASKED WHEN**

Maybe SSE3 is part of the hot spot problem?
**IF YOU HAVENT HEARD OF THE HOT SPOT PROBLEM THEN YOU HAVENT DONE ANY RESEARCH**

Maybe the deeper pipelines are contributing to the current leakage?
**SURE LOOKS LIKE THIS AND THE ABOVE WERE QUESTIONS NOT STATEMENTS**

Might it be better for Intel to current port the P4 to the 9 micron instead of trying to work out the problems with Prescott?
**IF THE ADDED FEATURES OF PRESCOTT ARE CAUSING THE NEW PROBLEMS WHY NOT**

Wignutz Now tell me one good piece of information you added to the thread?
NONE
YOU WENT ON THE ATTACK. IS THAT ALL YOU KNOW HOW TO DO?

 

[Platinum321]

Originally posted by: Ticktanium2038
I think I have made the comment enough times on buying a P4 2.4C enough that I cant be classified as Anti-Intel.
:: That really doesn't say much other than the fact that you are a price/performance shopper.

I dont see how prescott helps the Intel Mix ramp up their CPU line.
:: 64bit processing to compete w/ Hammer if deemed necessary.

Generally a next gen processor is a performance increase. I see prescott as nothing more that a slightly more expensive P4 that isnt an improvement over the existing P4 3.2.
:: In the past, this has been the situation. Some say that processor performance have come to the point where it's no longer as important of a factor as features and value. Centrino is a perfect example of non-performance oriented product, although it's performance is really nothing to laugh about.

I think sao123 had the best comment:
------
Well my first intuition is this... wait for the socket 775LGA prescotts. (BTW...this is when strained silicon is supposed to be introduced) They are supposed to debut Q3 04. This guarantees you arent buying a end of the roadmap mobo & cpu. IE...in the near future AGP will be replaced with PCIX and DDR II, also SATA is supposed to get a bump to 300MB by Q1 05 at the latest.
------

I think Intel releasing Prescott at existing speeds 2.8-3.4 will only make them look stupid when the benchmarks reveal the processor is slower than their existing line at higher speeds. Maybe waiting until the 775LGA where the core will be mature enough to qualify as a superior chip to thier own existing P4 line.
:: Speculations, this is what people are complaining about when they read what you have to say.

Why does someone want to spend more for something that perform less?
:: Good question, but that obviously will not be the case. People will pay more for value, and PSC will have value to justify it's cost whether it be performance or something else, otherwise it will not be released.

But I guess you couldnt keep your mouth shut wignutz and offer a good investor or technical perspective of the situation. Try answering the question next time instead of personal bashing because I called you out before. Still waiting for your 100+ watt cpu listing you claimed some time ago? Your personal bashing is really lame get a life.
:: So far, PSC will consume about 100+ watts.

 

[Wingznut]

*sigh*... I generally try to stay away from such dissect each others sentences posts, but I'll bite on this one.

But we'll need a couple of ground rules...
First off, links from theinquirer (or other such places) do not count as fact. Only reputable sources are valid.
Second, predicting performance from a pre-production stepping is inaccurate, as I've pointed out numorous times to you.
That being said, here goes:

It (Prescott) will not reach 3.6 until the second half of next year.
**TRY READING INTELS ROADMAP GENIUS ITS ALL THERE FOR YOU**
I'm not really a genius... You'll have to show me what you are talking about, as I have never seen said roadmap.

Its also said that overclocking a Prescott to 3.6 will cause it to burn up and die.
**YUP SOURCES CONFIRMED IT WILL BURN UP**
Again, I must be rather braindead, because I have yet to see any confirmation of the above. Happen to have any of these "sources" readily available?[/b]

A prescott at 3.4 will perform the same as a P4 at 3.2.
**TRY A WEB SEARCH PRESCOTT IS LOWER THAN P4 ITS OUT THERE**
Ok, I did a search. I found nothing that proves that a Prescott will be slower clock-for-clock than a Northwood. Again... Happen to have anything to back up your prediction?

The only thing prescott will add is SSE3...
**OK SO IT WILL ADD DEEPER PIPELINES AND BETTER BRANCH PREDICTION**
And strained silicon... And 1mb L2 cache... And better HT... And...

This (strained silicon wafers) would indeed help the leakage problem.
**NO KIDDING I DIDNT SAY IT WOULDNT I ASKED WHEN**You said that it "would indeed help the leakage problem"... How would SS wafers help current leakage. (Oh yeah, and what leakage are you talking about?)

Maybe SSE3 is part of the hot spot problem?
**IF YOU HAVENT HEARD OF THE HOT SPOT PROBLEM THEN YOU HAVENT DONE ANY RESEARCH**
Again, I must be a moron (in your eyes)... What "hot spot" problem is this? And how could a new instruction set contribute?

Maybe the deeper pipelines are contributing to the current leakage?
**SURE LOOKS LIKE THIS AND THE ABOVE WERE QUESTIONS NOT STATEMENTS**
I understood that, but you weren't asking if there was a current leakage issue... Even if there was/is, I don't understand how a deeper pipeline can affect that. Maybe you can enlighten me?

Might it be better for Intel to current port the P4 to the 9 micron instead of trying to work out the problems with Prescott?
**IF THE ADDED FEATURES OF PRESCOTT ARE CAUSING THE NEW PROBLEMS WHY NOT**
My point here was that you had already stated that "the only thing prescott will add is SSE3". So, if you really believed that, then wouldn't it already just be a "port"? (Just another example of you trying to spread misinformation.)

Wignutz Now tell me one good piece of information you added to the thread?
NONE
YOU WENT ON THE ATTACK. IS THAT ALL YOU KNOW HOW TO DO?
I could ask the same of you... What one GOOD piece of information have you contributed? It's nearly impossible to contribute to a thread that is so flawed from post #1.


Again... theinquirer is not a reliable source. Only reputable sources will be deemed valid proof for all that you say.

And I never answered your question about 100w cpu's because you twisted my words around. I told you that the discussion would be over if you continued to do so, and I stood by that. (Not to mention that others already provided examples, but you proceeded to ignore those.)

 

[Duvie]

I guess jack-ass didn't read this page, considering he responded in the thread in his normal foot-in-the-mouth I make shite up!!!! I thought there was hope for you butit is clear you arethe king of FUD...


Moderators please give tick (funny cause parasite comes to mind when I think of you) the title of FUD King!!!

Prescott benches!!!

look at super pi perfpormance at both 1m and 2m...near 3.2ghz performance...obviously helped by the added cache....OH YEAH, THE PRESCOTT ONLY HAS ADDED THE SSE3...MORON!@!!! Like you have never read an prescott thread in this place which is funny cause you spout more shite about it like you know something...

better in both 3dmark2k1 and 2003 then regular 2.8 northwood and I don't think they are quite sse3 optimised...NIce try though!!!

It does equal or better then the northwood 2.8ghz in every test except the cpubench 99 version while it does great in cpubench2003...I think that is a clear sign the program is so old it can't take advantage or identify the processor right...

Next page look at the none synthetic bench of aquamark..It did better then the northwood...So where is your proof..LOL!!!



Oh and finally it is only an early stepping and A0 would likely be the first and I have heard there have been 4 so I would think this is quite conservative...

 

[sao123]

intel roadmaps are generally available from...
theinquirer.net & xbit labs.

Here is the roadmap as it was 1/1/03
http://www.hwextreme.com/articles/prescott/images/roadmap_big.gif

here is the roadmap detailing the prescott delay.
http://www.theinquirer.net/?article=12040

 

[DivideBYZero]

Originally posted by: Ticktanium2038
It (Prescott) will not reach 3.6 until the second half of next year.
**TRY READING INTELS ROADMAP GENIUS ITS ALL THERE FOR YOU**

Its also said that overclocking a Prescott to 3.6 will cause it to burn up and die.
**YUP SOURCES CONFIRMED IT WILL BURN UP**

A prescott at 3.4 will perform the same as a P4 at 3.2.
**TRY A WEB SEARCH PRESCOTT IS LOWER THAN P4 ITS OUT THERE**

The only thing prescott will add is SSE3...
**OK SO IT WILL ADD DEEPER PIPELINES AND BETTER BRANCH PREDICTION**

This (strained silicon wafers) would indeed help the leakage problem.
**NO KIDDING I DIDNT SAY IT WOULDNT I ASKED WHEN**

Maybe SSE3 is part of the hot spot problem?
**IF YOU HAVENT HEARD OF THE HOT SPOT PROBLEM THEN YOU HAVENT DONE ANY RESEARCH**

Maybe the deeper pipelines are contributing to the current leakage?
**SURE LOOKS LIKE THIS AND THE ABOVE WERE QUESTIONS NOT STATEMENTS**

Might it be better for Intel to current port the P4 to the 9 micron instead of trying to work out the problems with Prescott?
**IF THE ADDED FEATURES OF PRESCOTT ARE CAUSING THE NEW PROBLEMS WHY NOT**

Wignutz Now tell me one good piece of information you added to the thread?
NONE
YOU WENT ON THE ATTACK. IS THAT ALL YOU KNOW HOW TO DO?

Yeah, I guess Wingz must just be asleep when he's at work. At Intel.

 

[bgeh]

Originally posted by: Duvie
I guess jack-ass didn't read this page, considering he responded in the thread in his normal foot-in-the-mouth I make shite up!!!! I thought there was hope for you butit is clear you arethe king of FUD...


Moderators please give tick (funny cause parasite comes to mind when I think of you) the title of FUD King!!!

Prescott benches!!!

look at super pi perfpormance at both 1m and 2m...near 3.2ghz performance...obviously helped by the added cache....OH YEAH, THE PRESCOTT ONLY HAS ADDED THE SSE3...MORON!@!!! Like you have never read an prescott thread in this place which is funny cause you spout more shite about it like you know something...

better in both 3dmark2k1 and 2003 then regular 2.8 northwood and I don't think they are quite sse3 optimised...NIce try though!!!

It does equal or better then the northwood 2.8ghz in every test except the cpubench 99 version while it does great in cpubench2003...I think that is a clear sign the program is so old it can't take advantage or identify the processor right...

Next page look at the none synthetic bench of aquamark..It did better then the northwood...So where is your proof..LOL!!!



Oh and finally it is only an early stepping and A0 would likely be the first and I have heard there have been 4 so I would think this is quite conservative...

and not to mention that most of these proggys have SSE3 support yet too
JMO

 

[alexruiz]

I was going to suggest an Athlon 64, but in answer to your question, pretzelcot WILL be faster than Northwood just by the larger cache. forget the rest of the improvements, just the cache will make it faster at the same clock frequency. Heat? Well, the ball in intel's engineer's court to deal with it. Right now, the leaked information shows it as a flamethrower, but this doesn't mean it wil be that way.

Regarding leaked numbers, it is performing very well (better than what AMD fans like me would like). Remember the slot A Athlon at stepping A0, or the hammer? They were beaten badly in the benchmarks. Come launch day and they made those prophets of doom eat their words.... So, this will very likely be the same case. How good?? Still unknown.