Fudzilla: Cypress yields 60-80% Fermi yields at 20% Fermi 20% faster than Cypress

Page 6 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
nitromullet

nitromullet

Diamond Member
Jan 7, 2004
9,031
36
91
lopri said:
It wasn't exactly small-die approach, iirc. G92 was designed for performance-mainstream part originally but due to the delay of GT200 NV had to retool it (with lots of VRMs) as a replacement for G80.

Obviously the so-called "Sweet-spot strategy" or "Small-die strategy" is a marketing talk. But I think there is some truth to it in that AMD designed its architecture to be fairly modular. Seeing the launches of Evergreen family, it looks like AMD can chop off the units as needed or even weave them together relatively easily. I think it is a smart approach.
Click to expand...

All true. IMO, the point here really is that since RV770 ATI has been hitting their design goals, whereas since G92 NV has been coming up a little short of their intended mark.

Going with a small die probably makes it a little bit easier to hit your intended goal (RV770), but the risk is that if you don't you have a much slower product (RV670 vs. G92). With a large die it's probably more difficult to get it right, but when you do you knock it out of the ballpark (G80 vs. R600). When you don't hit your marks with a large die you might still be able to compete, but at greater cost to you (GT200 vs. RV770).

I don't think there is an inherently better strategy, they both have proven to work when pulled off successfully. The trick is to pull it off, which is not a trivial matter. Everyone on the ATI side is looking for another RV770, and the NV side is hoping for another G80. That type of successful execution is rare, which is what makes them special.
 
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom