No, look at the title of the slide. And the bullet point for which that is the subtext:
Title: 65nm Volume Production ...
Bullet:
Seamless Production Transition
- Use of 90nm-proven, high-performance SOI transistor
- Next generation transistor in 3 months
Clearly it refers to the transition to 65nm, which AMD said was Q4, and which DailyTech has said is December, based on AMD OEM roadmaps.
And that production transistion will be seamless *because* they are using 90nm-proven transistor technology, and *then* 3 months later, they will move to a next generation transistor.
If they were talking about some 90nm transistor 3 months from NOW, it would not be in that bullet point, nor even on that slide. All IMO, of course.
Of course, AMD may well attempt to improve the 90nm transistors, they claim to try to do so once per quarter.
On the SiGe transistor performance increase, I believe it is claimed to be 1.4 / 1.24 = about 13% better than DSL.
(40% better than no strain, whereas DSL was 24% better than no strain)
---------
Actually, on reflection, I think they are trying to lower expectations for the first 65nm parts by emphasizing that the first transistors will be "90nm-proven", and saying 3 months later we'll use something better.
This fits with the low clocks roadmapped so far for the first 65nm parts (and Dirk's comments confirming this) |
