Comments
There's unfortunately not much to read here yet...
Best of Machine Learning
Discover the best guides, books, papers and news in Machine Learning, once per week.
Follow the full discussion on Reddit.
I have an old server that I'd like to convert to a machine learning testing platform. It has (2) Xeon E5-2670 CPUs for a total of 16 cores, 128Gb ram, and (2) PCIe x 16 slots. I bought a pair of Nvidia Tesla K80s but I am wondering if I would get better performance adding another pair of K80s since the prices are so good right now. The issue I have is that my motherboard only has (2) PCIe x16 slots. I was thinking that I could install (2) PCIe bifurcation cards, and add extension cables to plug-in all (4) K80s. If I understand correctly that would give each K80 8x8 PCIe lanes which would be further dropped to 4x4 for each GPU. Would the performance hit of the reduced lanes outweigh the additional GPUs?
There's unfortunately not much to read here yet...
Ever having issues keeping up with everything that's going on in Machine Learning? That's where we help. We're sending out a weekly digest, highlighting the Best of Machine Learning.
Best of Machine Learning
Discover the best guides, books, papers and news in Machine Learning, once per week.
Join 800 others in curating the best parts of Machine Learning!
Just one more step. We use social logins to make sure that real humans are curating the list!
or contribute on GitHub directly.