It might indicate the GM200/GM210 won't around for a while:
http://www.techpowerup.com/207265/nv...0-silicon.html
Edit!!
The Anandtech take on things:
http://www.anandtech.com/show/8729/n...-k80-gk210-gpu
Printable View
It might indicate the GM200/GM210 won't around for a while:
http://www.techpowerup.com/207265/nv...0-silicon.html
Edit!!
The Anandtech take on things:
http://www.anandtech.com/show/8729/n...-k80-gk210-gpu
One thing interesting about the GM204 that under constant computer loads it was not really a massive improvement in performance/watt over the previous generation,and the Tesla cards will be under such loads. So I suspect it made much more sense to just tweak the current GK110 instead.
Woah - passive cooling on a 300W TDP card? I know dense servers tend to have a lot of throughflowing air, but 8 of those in a node is going to need some monster airflow....
Interesting to see the cache redesign as well - wonder if the smaller caches are inherently limiting to performance, or if it's more specifically limiting to compute performance....?