If this is your first visit, be sure to
check out the FAQ by clicking the
link above. You may have to register
before you can post: click the register link above to proceed. To start viewing messages,
select the forum that you want to visit from the selection below.
Exciting News: Chaos acquires EvolveLAB = AI-Powered Design.
To learn more, please visit this page!
New! You can now log in to the forums with your chaos.com account as well as your forum account.
Depending on the price, could be a good option for people with limited PCI-e slots. The picture shows that the card consists of two GF110 fermi chips (the new fermi chip that debuted GTX 580). No telling how many CUDA cores each GF110 would have, but you'd figure it would have at least as many CUDA cores as the smallest fermi to date. Plus, having 3GB of dedicated RAM on one PCB would help with larger data sets, and one would assume that the power draw would be less than 2 separate 50% equivalent cards.
If this is real, could be a good card for GPU rendering, as long as nVidia didn't disable FP64 (double precision FLOPS) on this chip. They already disabled it on many of the multiprocessors of the GF100 series Geforce cards to separate it from the Tesla cards. The Quadro 2000 and 600 cards used newer GF104 & GF108 chips, but those were severely limited in terms of FP64 (something like 1/12 of FP32). The Quadro 4000/5000/6000 cards stuck with the GF100 chips because of the FP64 capability. Don't know how the GF110 stacks up in terms of being capable of FP64 computations, or if nVidia decided to disable it even if it is built to be capable of FP64.
Comment