AMD in 2006 ?

[Furen]

Current leakage, I believe. I dont think there's such a thing as volt leakage and I dont know what you mean by "speed leakage"...

 

[Viditor]

Originally posted by: Furen
Current leakage, I believe. I dont think there's such a thing as volt leakage and I dont know what you mean by "speed leakage"...

Exactly...current leakage.

Bona Fide, take a look at this excellent article Johan did here back in Feb, and you'll see what I'm talking about...

Johan's Article

 

[Bona Fide]

Originally posted by: Viditor

Originally posted by: Furen
Current leakage, I believe. I dont think there's such a thing as volt leakage and I dont know what you mean by "speed leakage"...

Exactly...current leakage.

Bona Fide, take a look at this excellent article Johan did here back in Feb, and you'll see what I'm talking about...

Johan's Article

Interesting...Johan seems to think Intel is going to crash and burn on the back of these multi-core processors (Pentium D) because of too much leakage. How I would love to see that 🙂

 

[sonoma1993]

i was reading some computer magazine one day at work, they had roadmap in there for both amd and intel for 2006. They show that amd was planning on coming out with a new m2 fx chip that will have 2mb of l2 cache. For intel they had a dual core chip coming out that was going to have 2mb l2 cache for each core. Im not sure how accruate the magazie was, if i remember right it was computer power user or somethen like that.

 

[Viditor]

Originally posted by: Bona Fide
[Interesting...Johan seems to think Intel is going to crash and burn on the back of these multi-core processors (Pentium D) because of too much leakage. How I would love to see that 🙂

I don't know about crash and burn, but I think it's evident that Intel hit a "leakage wall" much sooner than expected on the Netburst architecture (remember that they were originally predicting 10 GHz chips by next year)...
I've often wondered why Intel didn't go with SOI to keep the leakage down, and the only reason I can think of is that at the time of their decision SOI wafers weren't very plentiful. Certainly enough for a smaller company like AMD, but if Intel had to produce all of their chip line on SOI, I don't think there were nearly enough wafers available then...

Be that as it may, AMD will also hit a "leakage wall" at some point in the not too distant future, and IMHO the race is on to find a means of manufacture that vastly reduces leakage again.

Cheers!

 

[Bona Fide]

Originally posted by: Viditor

Originally posted by: Bona Fide
[Interesting...Johan seems to think Intel is going to crash and burn on the back of these multi-core processors (Pentium D) because of too much leakage. How I would love to see that 🙂

I don't know about crash and burn, but I think it's evident that Intel hit a "leakage wall" much sooner than expected on the Netburst architecture (remember that they were originally predicting 10 GHz chips by next year)...
I've often wondered why Intel didn't go with SOI to keep the leakage down, and the only reason I can think of is that at the time of their decision SOI wafers weren't very plentiful. Certainly enough for a smaller company like AMD, but if Intel had to produce all of their chip line on SOI, I don't think there were nearly enough wafers available then...

Be that as it may, AMD will also hit a "leakage wall" at some point in the not too distant future, and IMHO the race is on to find a means of manufacture that vastly reduces leakage again.

Cheers!

Too true, especially if AMD stays with SOI. For now, being a small(er) company, those wafers are enough for them. But I think AMD's is going to catch up and/or overtake Intel by the end of 2006...we'll have to see what new direction they go in then. 🙂