Hacker News new | ask | show | jobs
by pclmulqdq 380 days ago
CPUs are quietly becoming very well-balanced machines for BS 1 inference. The latest Intel Xeons should be at ~20 TPS.
1 comments
Spooky23 380 days ago
A base Mac Mini is ~20 :)
link
pclmulqdq 380 days ago
Oh yeah, I did that math not assuming any quantization. I think if you can get a 3-4 bit quant working + int8 math, ~80 might be achievable.
link