I’v found two MIO connectors on my Tesla C870 card, but i know that CUDA doesn’t suppost SLI mode. What these connectors are made for? For the future version’s of CUDA, or they needed in other Tesla’s solutions such as S870?
ps many thanks to nVidia, with Tesla solution we can compute our models and it take’s only several days against weeks on CPU-based computation
pps oh, sorry for my english, it doesn’t native for me