SLIDE algorithm

A recent algorithm developed by Rice University + Intel claims a ~3x speedup on CPU only as compared to Nvidia V100 and 10x speedup over Tensorflow CPU.

A lot of the details are over my head, so I won’t even attempt to describe it. (Not a very useful starting point. Apologies for that.)

Has anybody already looked into it and whether a Julia reimplementation would be useful given that is seems it would depart from the current focus on autodiff?


Hi, I am also quite keen on this topic. Any progress so far that you know of?
Please let me know.

Thanks & Best Regards

Hi Michael, as fair as I can tell nobody has tried implementing this algorithm in Julia. If you’re interested, I would recommend doing so and posting here once you have something workable :wink: