• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Mobile A64 cpu availability

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

SunTzu69

Member
Joined
Apr 30, 2004
Location
Canada
In 4-5 months, will we see any A64 Mobile (or DTR) 2800+, 3000+ or 3200+ on either the Clawhammer or Newcastle core ? Will they phase out the Clawhammer cores on these mobiles ? Or will they all be replaced with higher performance models like the desktop versions ?
 
According to The Wills of The Wind...erm, I mean, AMD's "official" timeline, there should be a new Athlon XP-M core built on 130nm silicon on insulator sometime in 2H04. New Mobile 64's will be Odessa cores...but those are going to be here--at the QUICKEST--in 2H04. Odessas are going to be 90nm SOI chips.

In other words, I don't even think AMD is sure when they will be ready.
 
Funny thing, everyone's been forecasting AMD's doom every since they existed. And just about 95% of the time, they've always outdone everyone's expectations. I don't know what AMD's got in store for us, but they usually know what they're doing. The Clawhammer is a pretty nice example of this. Just had to say it, carry on. :D
 
Gautam said:
Funny thing, everyone's been forecasting AMD's doom every since they existed. And just about 95% of the time, they've always outdone everyone's expectations. I don't know what AMD's got in store for us, but they usually know what they're doing. The Clawhammer is a pretty nice example of this. Just had to say it, carry on. :D

Where did I forecast doom?

I simply do not believe that AMD will stick us with something like Intel stuck us with Prescott. AMD cannot afford to lose the enthusiast (i.e. us) marketshare by handing us a faulty, overheating, 112-million-transistor-overheating behemoth requiring a 2.2lb copper brick for a heatsink.

90nm SOI will be delivered by AMD when it is ready which may be as soon as the start of 2nd half, 2004, or as late as December 2004.
 
Just a random thought I've been feeling like throwing out. ;)

AMD is getting quite difficult to predict, but I'd say they know what they're doing.

It would be logical for them to, in the near future, change their mobile parts over to the Newcastle core.

I also think that AMD will get .09 micron right, but it's a difficult barrier to break, and the on-die memory controller doesn't make things easier.
 
Gautam said:

...
I also think that AMD will get .09 micron right, but it's a difficult barrier to break, and the on-die memory controller doesn't make things easier.

Why on-die memory controller doesn't make things easier?
 
Bigger die size, and greater heat dissipation, I'd imagine. If you compare the Newcastle core to the Barton core, for example, you can see a significant increase in heat(89W for 3000+, 2.0GHz, 1.5v). The maximum rated voltage is much lower, at only 70 C. It doesn't, at first glance, appear to be an easy architecture to shift over.
 
The memory controller is more an architectural thing, than the power and heat issues which are related to transistor scaling, high frequency and leakage current problems facing the 90 nm technology, ....

At this point, I do not see memory controller being an issue for the 90 nm technology. Further memory controller itself and its I/O part are running at lower frequency, higher voltage, and that does not translate into the leakage current facing 90 nm SOI.

Higher voltage is needed for interfacing to the memory modules. Higher voltage does not mean higher leakage in this case, since the type of transistor used in the memory controller I/O are different, there are a lot more details here, .....
 
Last edited:
Gautam said:


AMD is getting quite difficult to predict, but I'd say they know what they're doing.
Have to disagree with you on that one! The only predictable thing about AMD is they are unpredictable - their strategy still seems to be in its formulation stage for the 64 bit line up. I have said before it would make operational sense to curtail clawhammer production for a number of reasons - however it could be that for other reasons (sales and marketing related) they decide to keep the clawhammer alive.
The lack of any clear information from AMD regarding what line up they will be releasing on 939 launch day also suggests an incomplete marketing strategy.
As for mobiles and possibly desktops in the future - there may be 256KB L2 versions as well as well as the current (shown on AMD specifications) 2700 and 2800 Newcastles, and the 2800, 3000, 3200 and 3400 Clawhammers
 
Back