Scotty OCs to 4.2Ghz on Air

[Mephistopheles]

So apparently, Prescott can overclock to 4.2Ghz on air cooling... Not bad for a processor with alledged heat problems, eh?

<A HREF="http://www.vrforums.com/showthread.php?threadid=12686" target="_new">Here is the link,</A> but it loads extremely slowly.

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

 

[falecomcyber]

i'm scared!
i'm 56k, hehe :-D

but, i'l take courage to view this!



I'm from Brazil/PE/Recife

 

[Mephistopheles]

It's pathetically slow.

I just finished reading the first page.
Trying... to get to... the second... page... ugh.


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

 

[Mephistopheles]

Wait a second... You're from Brazil?

I'm from Campinas/SP/Brazil...

<i>Então, porque todo o inglês? Olha um brasileiro aí! "Fale com cyber?"</i>

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

 

[pitsi]

As far as I am concerned, this is very weird. Especially when having in mind, that all roadmaps indicate nothing faster from Intel than 3.8GHz for the whole 2004. If it is so easy for them to get 4.2GHz out of Prescott, then what's all this fuss with Prescott's heat problems, etc.

 

[HardWareBoss]

That was with the first batch of CPUS

<P ID="edit"><FONT SIZE=-1><EM>Edited by HardwareBoss on 02/01/04 11:42 AM.</EM></FONT></P>

 

[Mephistopheles]

No, actually, Intel has changed that. They will be offering a 4Ghz Prescott still in 2004. Check <A HREF="http://www.theinquirer.net/?article=13886" target="_new">this out.</A> And this guy was able to clock prescott at 4.2Ghz with a current (or maybe a bit old, who knows) stepping, meaning that by the end of the year, they could still work on scaling Scotty to speeds above 4Ghz, if they needed to.

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

 

[HardWareBoss]

VR Zone all ways takes forever to load. I have cable.

 

[Vapor]

Good thing it's only two pages...still waiting for the first page to load though.

Maxtor disgraces the six letters that make Matrox.

 

[Vapor]

I'd like to see one with a Prometeia...or another LN2 experiment. Benchies tomorrow!

Maxtor disgraces the six letters that make Matrox.

 

[falecomcyber]

HEahuehau

Loko encontrar alguem do brasil aki...
vc frequenta outros forums (em portugues) ??

:-D

I'm from Brazil/PE/Recife

 

[Mephistopheles]

Ih... nem freqüento, não... Nada em português... hardware, no Brasil, é uma tristeza... Não tem nada bom! Se conhecer algum, me avisa...

Mas eu navego bastante... e uso messengers da vida... esse tipo de coisa...

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

 

[Kelledin]

Hmmm...unfortunately, it looks like Visionary just got lucky. Or maybe he's doing something different with his sample.

Most other review sites point out that Prescott is running about 10-20C hotter clock-for-clock than Northwood. The few who did OC testing got Prescott to maybe 3.6-3.8GHz. <A HREF="http://www.hothardware.com/hh_files/CCAM/p4prescott_p4ee34(3).shtml" target="_new">HotHardware</A> in particular was able to get better OCs out of Northwood.

<i><Lionel Hutz> I'll be defending...The SCO Group!!!??? Even if I lose, I'll be famous!</i>

 

[Mephistopheles]

Are you really sure?

An effortless overclock gave us 3.72GHz; we could POST at 4GHz but we didn’t want to showcase what was ultimately possible with Prescott, rather what was easily attainable without increasing voltages.

Intel could have launched Prescott at higher clock speeds than they did, however it seems that their desire to produce as many mainstream Prescotts as possible (2.80E in particular) won out in this case.

Think again. Check <A HREF="http://www.anandtech.com/cpu/showdoc.html?i=1956" target="_new">anandtech's review</A> for more info... This is a great review.

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

 

[Kelledin]

Yup, pretty sure.

<A HREF="http://216.239.37.104/translate_c?hl=en&ie=UTF-8&oe=UTF-8&langpair=de|en&u=http://www.hardtecs4u.com/reviews/2004/intel_prescott/index11.php&prev=/language_tools" target="_new">- Click -</A> HardTecs4U

With all then following steps we had observed the temperature development in each case by means of Main board monitor. In the not over-clocked condition we determined 55°C core temperature under load. We made ourselves and selected only once without tension addition to the work a FSB of 213 MHz, which corresponded with the multiplicator of 16 3,407 MHz to CPU clock. This accepted the processor also without troubles and went through stably the stress tests. We could not register a new temperature peak value. It remained with 55°C, which corresponded to the not over-clocked condition. Also in the next step with 16 x 218 MHz, thus 3488 MHz CPU clock, also here required it still no spannungserhoehung and it remained under load with 55°C.

A jump on 225 MHz FSB (3600 MHz CPU clock), failed then however. We increased the tension therefore to 1,4V without success. Afterwards we selected 1.425V, which was varying indicated us after MBM with 1.41-1,43V. Material actually only 1,409V tension fitted. We however already did not obtain stability also here, thereby after quite short time stress (approx. 5 minutes), up to the program crash, a core temperature of approx. 58°C. From that to make in the chapter before shown differences at the temperature temperature between the ASUS and Intelmainboard, we refrained then from it, further spannungserhoehungen.

Stably we could 223 MHz FSB and/or 3570 MHz CPU clock, with lying close 1.409V and a "core temperature" (as per MBM on Pentium 4P800) from 59°C under load reach. This speaks in principle, despite some imponderables, for which possibilities are already with the momentary manufacturing still in the Prescott, in addition, but that further clock paste run gene either or clearly better cooling concepts needs a better manufacturing process.

Obviously it's still a bit of a mixed bag. It also speaks volumes that in the middle of rushing the market towards Prescott <b>ASAP,</b> Intel is forced to trot out another Northwood iteration to actually reach 3.4GHz.

Prescott needs to at least be clockable (thus <i>over</i>clockable) beyond Northwood. After all, it's what the core is designed for. Hell, it's what Intel sacrificed performance and efficiency for in this case.

Hopefully clock speed gets better once yield improves. Otherwise it's a lot of engineering sacrifices for nothing.

<i><Lionel Hutz> I'll be defending...The SCO Group!!!??? Even if I lose, I'll be famous!</i><P ID="edit"><FONT SIZE=-1><EM>Edited by kelledin on 02/01/04 06:06 PM.</EM></FONT></P>

 

[Mephistopheles]

I'm not really completely convinced yet...

If anandtech said it's easy to overclock, and a number of other sites also claim to have achieved ~3.8Ghz easily, then, well, isn't it possible that that particular sample HardTecs4U got was a worse one? Consider the ease with which anandtech got to 3.8Ghz, and you'll see what I mean. Unless anandtech is lying, of course.

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

 

[Mephistopheles]

BTW, I could read that whole article in its original language - I speak German.

It's a great conclusion...

It reads something like this:

<my translation>
Those who were expecting Prescott to set new records in performance must be disappointed by now. Keeping things in perspective, however, Prescott does introduce better Hyperthreading and its other improvements, and it turns out to be some 7% faster than Northwood at the same clock speeds in some applications. On others, though, it's only an equal to northwood, and on a few select others, Northwood can be faster (like 12% for Lightwave or Comanche).

What must still be said is that Prescott definitely scales better than Northwood; our higher-clocked prescotts compared better to equivalent northwoods than the lower-clocked ones. This can truly make Prescott's performance better, and it also means that Intel is preparing to ramp the speed of this processor.

In any case, with Prescott, Intel Processors are still performing better than equivalent AXPs, but still performing worse than equivalent A64s. But, because the prices for both A64 and prescott are about the same, each buyer has to take a pick based on each processor's strengths - the P4 outpaces A64 in a number of benchmarks. So each buyer buys what he/she wants based on individual application performance.
</my translation>

It's something like that, loosely translated.

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

 

[Terracide]

I like this quote from Anandtech:
"To put it bluntly: Prescott becomes interesting after 3.6GHz; in other words, after it has completely left Northwood’s clock speeds behind".

I think that says it all...


Don't pretend - BE!

 

[pIII_Man]

Toms...although now somewhat unreliable said that they did not do the overclocking test because they couldnt pull much out of the scotty...

Also i have to submit that i have started hearing reports of over 3.8ghz on air cooling with some of the newer batches of NW 3.0's


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

 

[Terracide]

Another benchmark test:
<A HREF="http://www.xbitlabs.com/articles/cpu/display/prescott-tests_3.html" target="_new"> Xbitlabs </A>

Don't pretend - BE!