Possible new stepping for Winchesters in January?

Quote:

---

Originally Posted by OC Detective

AMD's SSDOI technology on 90nm will differ from that currently used on the FX55's. The SSDOI used at present is similar (read identical!) to that used by IBM however for 90nm (second generation) it will differ as it will be localised to certain components and not involve the removal of silicon germanium.

---

So does that mean the new one will be not as good? Is it prehaps cheaper maybe?

Well if you listen to AMD's Tom Sonderman it will be better (second generation typically means that!).

Quote:

---

Originally Posted by OC Detective

AMD's SSDOI technology on 90nm will differ from that currently used on the FX55's. The SSDOI used at present is similar (read identical!) to that used by IBM however for 90nm (second generation) it will differ as it will be localised to certain components and not involve the removal of silicon germanium.

---

This is true. BUT...

IBM has already anounced it plans on doing the same thing in 1Q 2005 in order to get past the leakage problems it is facing with its high end PowerPC chips. That means that both the AMD and IBM processes will again be essentially the same. This is precisely why both AMD and IBM are announcing "good things" to come early in 2005. If you read some recent Apple press releases you'll see that Apple has become rather bullish about releasing its 3ghz+ PowerPC powered rigs right about the same time AMD is expected to pump out its high end 90 nano chips.

I hate to disappoint the fanboys, but IBM and AMD are going to be in bed together on this. Again, that's not a bad thing. AMD's R&D is focused on chip design, not process engineering and fabbing. And its the design that wins the day.

BTW, no ego battles here. I've done nothing but site common media references to what the facts are so that misinformation is not spread. Folks can beleive whomever they wish:)

Quote:

---

Originally Posted by OC Detective

Well if you listen to AMD's Tom Sonderman it will be better (second generation typically means that!).

---

Just wondering cause not all new revisions of things work out better, especially when it comes to technology

Don't fear, this will be a good thing. Remember, this isn't totally new technology. Its a refinement. Remember when AMD added a layer of silicon to the Thoroughbred? It did wonders.

Quote:

---

Originally Posted by Super strokey

i thought you were still waiting to get the winnie? If not maybe you should upgrade the vidcard first ;)

---

I own a 3000+ winnie, waiting for my dual channel ram to come! the mails slow I hope rather than got lost :stick: also im really hoping the ram works dual channel like it did for him. also im hoping winchesters work fine on nf3 mobos cause some say winchesters are too new for nf3 and you need a nf4 for 100% stablilty.


"Geforce4ti4200
Yes these chips will go 3 GHz on air. Happy?"


I was wrong, the first batch of 90nm wont do 3GHz but the 2nd batch will, heck itll soon be stock 3GHz after 2.6 and 2.8 are initially released. Two things I wonder: will those venice work with nf3 or are they too new and need a nf4? secondly will amd make lower speeds such as 3500+ when/if?

Hey uh.. call me dumb, but what is venice? I'm assuming a new chip, but I havn't heard of it. :(

Venice = rev E Winchester probably with SSDOI.

So they are dropping the Winchester name then? It's still the same thing, they are just improving a few little things, mostly the memory controller.

Quote:

---

Originally Posted by mdzcpa

This is true. BUT...

IBM has already anounced it plans on doing the same thing in 1Q 2005 in order to get past the leakage problems it is facing with its high end PowerPC chips. That means that both the AMD and IBM processes will again be essentially the same.

---

No argument that AMD shared IBM's strained silicon technology which for the FX55 is presumably much more beneficial to it than the PowerPC because the FX is still at 130nm. However I must admit I was not aware that IBM had also announced an iteration to their existing process - make sense though given the yield problems they had - dont suppose you have a link regarding it do you? Tried googling IBM strained silicon but to no avail!
Edit actually when I think about it SSDOI is probably not the correct terminology to describe the second generation of strained silicon application.

I dont know if venice will work on existing nf3 mobos, even if so, the lowest one may be 2.6GHz so it could be 2 years before those drop to $200 anyway

well... I'm under the impression 'the winchester is getting a new stepping' but I am also told that this winchester with the new stepping will not be a winchester anymore, but a venice. Which is not normally how things work... If the winchester is simply getting a new stepping, then they will be the same as the current one (3000+ 3200+ and 3500+) and more down the road.

However, if these new stepping will be 'venice.' Then the winchester is not 'getting a new stepping' and the winchester will remain with the problems it has always had.

Guys dont get too hung up about code names - these are only internal names AMD uses and not names to represent the core per se just a name they use when there is a distinct iteration to the process. Venice is just rev E Winchester possibly with localised strained silicon, Anandtech have been taking about this for well over a month. Where G4Ti4200 thinks that only 2.6Ghz and up will be Venice I have no idea!
http://www.anandtech.com/mb/showdoc.aspx?i=2264&p=3