Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
aurareturn
21 days ago
|
parent
|
context
|
favorite
| on:
Cerebras launches Qwen3-235B, achieving 1.5k token...
Probably because they are loading the entire model into SRAM. Thats how they can achieve 1.5k tokens/s.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
