Hacker News
new
past
comments
ask
show
jobs
points
by
hamiltont
1 days ago
|
comments
by
peri-cl
1 days ago
|
next
[-]
I think M1 through M3 were compute bottlenecked in prompt processing (hence the very large gap between M3 and M5, in this page's benchmarks, that's not explained by memory bandwidth alone).
For
generation
speed in isolation, yes.
reply
by
GeekyBear
1 days ago
|
parent
|
[-]
The M5 generation added tensor instructions to the GPU cores.
reply
by
Lwerewolf
1 days ago
|
prev
|
[-]
This one is 2x m5 max, so ~1.2TB/sec.
reply
by
1 days ago
|
parent
|
[-]
deleted
reply