Prescott slower than Northwood "C"?

[Spitfire_x86]

Seems so from Anandtech Radeon 9800XT review. They used Prescott 2.8 GHz for their benchmarks.

In UT2003 Flyby test, it pulled 212 fps with Radeon 9800 Pro 256 MB. With the same card and same settings, P4 3.0C and P4 3.2C scored 226.3 fps and 232.8 fps respectively. If Prescott scales like Northwood "C" in this benchmark, then it should do 218.5 fps @ 3.0 GHz and 226.1 fps @ 3.2 GHz, with is slower than equally clocked Northwood "C". But it's 1 MB L2 cache should offer some performance boost over Northwood "C" in this benchmark if everything else is equal to Northwood "C".

One thing is certain- Athlon 64 isn't losing it's gaming performance crown in this year, and Athlon 64 3200+ should beat (or tie) early Prescott in gaming benchmarks.

----------------
<b><A HREF="http://geocities.com/spitfire_x86" target="_new">My Website</A></b>

<b><A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new">My Rig & 3DMark score</A></b>

 

[pIII_Man]

i beleive coolsquirtle posted a link to precott benchies where a 2.8c beat a 2.8ghz prescott in many benchies.


If it isn't a P6 then it isn't a procesor
110% BX fanboy

 

[Spitfire_x86]

This is from Anandtech. Coolsquirtle's link was from a Taiwanese website.

----------------
<b><A HREF="http://geocities.com/spitfire_x86" target="_new">My Website</A></b>

<b><A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new">My Rig & 3DMark score</A></b>

 

[Mephistopheles]

Noone said that was a prescott... the inquirer just mentioned the possibility...

<font color=red><b>M</b></font color=red>ephistopheles

 

[Spitfire_x86]

Prescott was written in white text

<A HREF="http://forumz.tomshardware.com/hardware/modules.php?name=Forums&file=viewtopic&p=371355#371355" target="_new">http://forumz.tomshardware.com/hardware/modules.php?name=Forums&file=viewtopic&p=371355#371355</A>

----------------
<b><A HREF="http://geocities.com/spitfire_x86" target="_new">My Website</A></b>

<b><A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new">My Rig & 3DMark score</A></b>

 

[guesswhosback]

prescott has gotsr to be just crushing u

its overheated
late
and slow
and incompatible

come on over to the performance crowd

the AMD crowd

since u claim to have no preference

 

[spud]

Its a NDA breach we have been through this before.

Discussing this is very moot since Anandtech claims it was a type'o. Also the Asian site that claimed they had a Prescott were also in correct according to my contact at Intel.

This is anti-hype at work, usually only occurs when AMD is trying to gain some inches on the global totem pole, I wouldnt even sweat it.

~Jeremy
Unofficial Intel PR Spokesman.(nVidia fill in rep for CoolSquirtle)

<A HREF="http://forumz.tomshardware.com/modules.php?name=Forums&file=faq&notfound=1&code=1" target="_new">Busting Sh@t Up In My Buddies Face!!!</A>

 

[HardWareBoss]

Still waiting on Real Benchmarks!!!! On Scotty.

 

[simwiz2]

LOL a typo! Yes, they accidentally typed "Prescott" with the system specs. And they accidentally made it the same color as the background of the page. Interesting.

 

[spud]

Thats what they say, just repeating what I was told.

~Jeremy
Unofficial Intel PR Spokesman. (nVidia fill in rep for CoolSquirtle)

<A HREF="http://forumz.tomshardware.com/modules.php?name=Forums&file=faq&notfound=1&code=1" target="_new">Busting Sh@t Up In My Buddies Face!!!</A>

 

[eden]

Spit, your calculations are far off man.

A 3GHZ is 7.1% higher clocked over the 2.8. Even in a bad case scenario of 5% performance scaling, you still get at least 222FPS.

At 7% you get an exact 3GHZ C 226FPS.

Anyways even if this was a Prescott, it could be an old sample, as Intel is likely fine tuning the current for competition, and these early mainboard chipset drivers are far from being adequate.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol:

 

[pitsi]

I highly doubt though that Prescott will be 20% faster than Northwood, clock-to-clock, as some reported in this forum a few weeks ago.

Also, if AMD manages to release Athlon 64 FX-53 (939 pin) the same time Intel makes Prescott widely available in Q1 2004, I believe it will be a very close battle. Some (myself included) thought that Intel, had something ready after Hammer got delayed for a whole year and simply waited for some competition to release it, but this has turned out to be soooo wrong!

 

[spud]

Its yet to be seen yet, dont be jumping to conclusions it could be technical it could also be mass production capacity isnt ramped up yet. Could even be the economics of the thing. We dont know for sure and we shouldnt overly specualte since we all did that with the hammer and now look at the chip.

~Jeremy
Unofficial Intel PR Spokesman.(nVidia fill in rep for CoolSquirtle)


<A HREF="http://forumz.tomshardware.com/modules.php?name=Forums&file=faq&notfound=1&code=1" target="_new">Busting Sh@t Up In My Buddies Face!!!</A>

 

[Mephistopheles]

prescott has gotsr to be just crushing u

its overheated
late
and slow
and incompatible

come on over to the performance crowd

the AMD crowd

since u claim to have no preference

There's nothing crushing me.

Also, your last few remarks seem to want me to become biased. Why would I want that?

I was merely stating that it's too early to jump to conclusions.


<font color=red><b>M</b></font color=red>ephistopheles

 

[HardWareBoss]

Why are they using 2.8 Scotty when the first release is 3.2 or 3.4. 2.8 1st qtr 2004. Big deal.

 

[Mephistopheles]

I was just thinking... If the "EE" line goes on, we'll be seeing Scotty with 2MB L3 cache (3MB total)... now that would be quite a processor...

I wonder if Scotty is as bad as it looks right now. I think we have to wait a bit and see for ourselves when truly comprehensive, comparative reviews are out. *oops, already said that*

<font color=red><b>M</b></font color=red>ephistopheles

 

[eden]

A Prescott with 2MB L3 will be wasted. Consider that the L2 is much faster than L3, they can still stand to add more L2 before latency becomes an issue. IMO they should reduce L3 to 1MB and put it. It'll likely still perform just as good.

Besides, aside from the fact P4s get a lot from cache increases, you have to wonder if those Prescott 2.8s even had the advancements. One would expect a lot more with PNI, improved HT, L1 cache increase as well as Trace Cache size increased!

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol:

 

[HardWareBoss]

Scotty L2 Cache 1MB, Does it need L3 cache 2MB?

 

[Spitfire_x86]

I've calculated in this way.

In case of Northwood "C", 3.25% performance increase for 6.67% clock speed increase. So there is 0.4875% performance increase for 1% clock speed increase

Anyways even if this was a Prescott, it could be an old sample, as Intel is likely fine tuning the current for competition, and these early mainboard chipset drivers are far from being adequate.

Since everybody is claiming that Prescott is not going to use a new mobo/chipset, then why early chipset drivers can cause problem?

----------------
<b><A HREF="http://geocities.com/spitfire_x86" target="_new">My Website</A></b>

<b><A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new">My Rig & 3DMark score</A></b>

 

[eden]

It can fit in all it wants, but the revisions and BIOS' may not be tuned for it.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol:

 

[Mephistopheles]

The hardware may just be in an experimental stage... Remember those very early A64 benchmarks, with much-lower-clocked A64s like a 1.6Ghz? They didn't look like a big deal at all...

<font color=red><b>M</b></font color=red>ephistopheles

 

[Spitfire_x86]

Final Athlon 64 performance varied little from 1.6 GHz preview. From 1.6 to 2.0 GHz, there's 25% clock speed improvement and A64 scales very well. So it looks much better than 1.6 GHz preview.

Moreover, 1.6 GHz A64 was labeled 2800+ and current 2.0 GHz A64 is labeled 3200+, 25% clock speed increase for 15% PR increase. This is another reason why it looks much better than preview.

----------------
<b><A HREF="http://geocities.com/spitfire_x86" target="_new">My Website</A></b>

<b><A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new">My Rig & 3DMark score</A></b>

 

[eden]

I'll admit one thing, it's that the PR rating is fairly well on track. In fact, for a 200MHZ jump next time, I expect a 200PR increase. That would actually make it conservative and extremely powerful.
Still, it COULD be better. But at least it's no stink-lying PR like the XP3200+ was.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: