Scotty E0 stepping: 4.2Ghz on air

[Mephistopheles]

...according to <A HREF="http://babelfish.altavista.com/babelfish/trurl_pagecontent?lp=ru_en&url=http://www.overclockers.ru/news/newsitem.shtml?category=1&id=1091282990" target="_new">this (babelfished) link</A>, prescott on E0 stepping @ 4.2Ghz and with 1200Mhz FSB is doable on air, albeit on an engineering sample. Looks as if, with a further revision, they could tackle 4+ Ghz speeds indeed (2005).

=<speculation>=============

I wonder if it would be too difficult for them to surprise everyone with a 1333Mhz FSB instead of a 1066Mhz one in 925XE? To ramp up DDR2-667, which is already available in timings equal to DDR2-533 (better than JEDEC specs)?... It would be great news, and would probably make a great difference in performance! And push DDR2/LGA775 much, much faster: DDR2-667 at current 4-4-4 or so timings in dual-channel in sync with 1333Mhz FSB would outperform any DDR1 config possible!

As for doing it unexpectedly, they did it once when moving from 533->667Mhz FSB; they jumped 667 and went for 800Mhz (that's a 50% increase in FSB)... And we all know what kind of performance increases Intel got then. It might even help socket-compatibility with Itanium, because itanium's new 667Mhz FSB (128bit - due next year, probably) has the same data bandwidth as a hypothetical 1333Mhz FSB... Too bad they're in trouble... So the likelyhood of them actually surprising us is about... 2.3% right now...

Feeding the execution core with more data from memory is probably a great idea. And a CPU-memory interface of 10.6GB/s would be more than enough to brag about...

It's probably more achievable than going 4Ghz this year. Multiplier 10, 3.33Ghz, could be a great prescott processor... Multiplier 11, 3.67Ghz, would make everyone respect prescott a little...

Just pure speculation, of course. It's just that an increase to 1066Mhz probably won't herald the kind of performance increases required in order to really stay in the game with AMD, if AMD sticks to their roadmap...

=</speculation>============

<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><P ID="edit"><FONT SIZE=-1><EM>Edited by Mephistopheles on 08/01/04 02:41 PM.</EM></FONT></P>

 

[coylter]

Great news !!!
Maybe intel will have some good things by chrismas (the point where i'll upgrade) !!!

Athlon 2700xp+ (oc: 3200xp+ with 200fsb)
Radeon 9800pro (oc: 410/360)
1024mb pc3200 (5-3-3-2)
Asus A7N8X-X

 

[Mephistopheles]

They probably will, but how will those stack up against AMD's offers? I wonder...

The only actual news is that the E0 stepping in prescott should enable reasonable yields and volume at 3.6Ghz and possibly 3.8Ghz, while the 4Ghz isn't impossible 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>

 

[trooper11]

well ill wait and see actauly non engineering sample performance, but if thats true then it would seem intel finally worked out some kinks in prescott to get it to be viable at all, itll be intersting to see the temps.

one thing i would say about jumping 1066fsb is that while yes there is ddr2 667 memory out there, its in evne less supply then ddr2 533 and im sure youve read the stories about memory manufacturers reporting slow adoption and less then expected sales so far. The adoption of ddr2 will be slow i suspect, not until either intel makes a more affordable solution , for example using agp instead of pci-e, or amd joins in on ddr2. as it is, a person wanting ddr2 has to also buy a pci-e card, which are nearly non existant right now. making so many new advancements in one solution does limit adoption more hten if it was just one part that had changed. so you could say pci-e cards are holding back ddr2.

i dont doubt it would be a performance boost, im just saying froma sales perspective, intel wouldnt be making anyhting extra at this point by doing so, just push back adoption even more. And intel wouldnt do somehting just for the enthusiasts, they show agian and again they care very little, if at all, about that segment. i dont really see ddr2 becoming viable until H1 '05, thats whne most memory manufacturers expect to ramp up production of ddr2 over ddr.

now amd adopting ddr2 would make adoption quicker, but i dont think thatll happen anytime soon either, it will have pci-e before it has ddr2.

 

[Mephistopheles]

now amd adopting ddr2 would make adoption quicker, but i dont think thatll happen anytime soon either, it will have pci-e before it has ddr2.

I don't really think AMD moves so much market share around as of now. It has been improving because of their technical lead at the moment, but I don't think AMD would truly change the pace of DDR2 adoption by adhering... Could be wrong, though.

I agree, however, to your general point. It would be unpractical to skip 1066Mhz altogether, but remember, that's what they did when going to 800Mhz FSB. Everyone expected 667Mhz to be paired with dual DDR333, and DDR400 was a risky and bold move, because DDR400 was hardly polished...

Personally, I think it's obviously unlikely that they'd skip 1066Mhz, but who knows?... It'd give them the performance boost they need... What I also hope to see is a decent 64-bit implementation in EM64T, not some crap. We'll see how EM64T stacks up against AMD64 soon, hopefully.

<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>

 

[phial]

Wow thats awesome. on AIR, haha..


looks like there is a method to their madness

-------
<A HREF="http://www.albinoblacksheep.com/flash/you.html" target="_new">please dont click here! </A>
<A HREF="http://www.subhi.com/keyboard.jpg" target="_new">This is you, interweb junky</A>

 

[ChipDeath]

gotta love babelfish:

One should immediately note that the acceleration was conducted on the modified maternal pay Asus P4c800- e deluxe with the use of air cooling. The model of carmelized sugar is not indicated,

---
Epox 8RDA+ V1.1 w/ Custom NB HS
XP1700+ @200x10 (~2Ghz), 1.4 Vcore
2x256Mb Corsair PC3200LL/1x512Mb Corsair XMS PC4000 2.5-3-3-7
Sapphire 9800Pro @412/740

 

[Mephistopheles]

I found that hilarious too...

Carmelized sugar? WTF has that got to do with anything???? :lol:

And what was that "maternal" talk? I wonder...

<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>

 

[trooper11]

well its not just amd adopting it that would help, but if amd adopted it and intel made it available on all of its line, meaning lower priced systems, adoption would be faster. i look back at the ddr beginings and how amds push did end up pushing intel to change over completely when rdram sells plumeted

 

[Mephistopheles]

True, but it was Intel that pushed DDR400 massively. Some even criticized that because it wasn't quite finished. Some speculated that DDR400 would never take off. Yet it did.

Sales on the i875 and i865 went well, and that made the memory industry develop DDR400 to be what we currently know as DDR400: a highly optimized product, with low timings, and high performance. That's the apex of the DDR1 memory; and that apex, if it ever comes for DDR2, is either on DDR2-667 or DDR2-800. In any case, Intel has already said that DDR2-667 might be supported in i925XE, and I don't think they'll refrain from supporting it after both SiS and ATi have announced plans to support it.

So DDR2-667 will probably get support. But since 1066Mhz FSB is already delayed, couldn't they just skip it and bring out a spanking new 1333Mhz FSB by christmas?...

<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>

 

[endyen]

A few points you seem to have missed.
Amd may not have much market share, but that share is more enthusiast based, so may make the demand for high speed ram happen a lot faster.
Intel did not "jump" the 666 fsb out of choice, but rather out of fear of the bible belt. They are already seen to be too close to "the beast" without addopting it's number.
I know you want Intel to shine, and they will, it just takes time.

 

[coylter]

LOL they could have just said 667

Athlon 2700xp+ (oc: 3200xp+ with 200fsb)
Radeon 9800pro (oc: 410/360)
1024mb pc3200 (5-3-3-2)
Asus A7N8X-X

 

[endyen]

Everyone would have known.

 

[Johanthegnarler]

Lmao. I love it.

<A HREF="http://arc.aquamark3.com/arc/arc_view.php?run=277124623" target="_new">http://arc.aquamark3.com/arc/arc_view.php?run=277124623</A>
46,510 , movin on up. 48k new goal. Maybe not.. :/

 

[darko21]

Re: I know you want Intel to shine, and they will, it just takes time.

Exactly Intel is going nowhere amd does not have capacity to take to much market (ATM)... Intel is known for making reliable fast products the many recent fumblings is only cause they cannot accept they are not the performance leader (ATM)... We saw this play out before with the introduction of the athlon and Intel rushed premature chips to market but not quite to the dergree now. Big companies who are known as the leaders have a great deal of trouble acepting loss of the crown. Look at nVidia's recent woes. nVidia fumbled over and over but it looks like they got the crown back now.

My personal prediction is intel will lag amd in performance for a year or 2. If fab 36 gets online and k9 is a winner AMD might grab some market share from Intel. I personaly think this would be good for the consumer. Competition is good and amd's share is small, Intel nearly had them in chapter 11 just over a year ago. So don't get too upset at the thought of AMD grabbing some market share it's probably a good thing over all for any consumer or would kanavite rather have paid 700 dollars for his 3.0 giz prescott cause I'd bet that's what it would cost if amd had gone chapter 11. Intel will get the crown back its just a matter of when and for how long.


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.