Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
nextaccountic
9 months ago
|
parent
|
context
|
favorite
| on:
Intel x86 documentation has more pages than the 65...
With FP16 you can fit twice as much weights in cache, and also fetch twice as much weights from memory
Also this depends on the size of the matrix
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
Also this depends on the size of the matrix