3 ms·
I'm not a pro in this field but I would think with AMD acquiring Xilinx they see a better way to the AI competition through FPGA. I would suspect others are in
by chrismarlow9 3y ago
I'm not a pro in this field but I would think with AMD acquiring Xilinx they see a better way to the AI competition through FPGA. I would suspect others are in the same boat. GPU for AI is a nice hack, but still a hack. Make hardware that specifically is for that job.
- tverbeure 3y agoHow can a notoriously area inefficient technology like FPGAs be hardware that is made specifically for the AI job? And I’m not even touching the fact that FPGAs (which I love to work with!) are a pain to program and optimize for.
- chrismarlow9 3y agoI mean I'm not that well versed on it but a quick Google search kinda points to faang, Tesla, and others investing in FPGA for AI things.
- tverbeure 3y agoMicrosoft has historically been the biggest proponent for using FPGAs for AI. It’s not that you can’t use them, and there may be narrow cases where they have a benefit. There are always trade-offs to be made. But that doesn’t make them the best hardware for AI (there’s silicon overhead that will never be used) and it doesn’t change the fact that they’re much harder and slower than bare metal CUDA to optimize for, in a fast moving field. Look at it this way: if you have a fixed DL architecture and a fixed set of network weights, there’s no question that you could come up with an ASIC that only does that, that will be cheaper than an FPGA or GPU, faster, and use lower power too. It will be the perfect inference solution for that specific problem. It would also be obsolete in a few months, before the chip comes back from the fab.
- scns 3y agoOff topic but RME uses FPGAs in their audio interfaces and delivers sub Firewire lantencies over USB2.
- DeathArrow 3y ago> Make hardware that specifically is for that job. That means an ASIC.