Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
ec109685
on Feb 4, 2023
|
parent
|
context
|
favorite
| on:
Open Assistant: Conversational AI for Everyone
Another idea is to dedicate cpu cycles to something else that is easier to distribute, and then use the proceeds for massive amounts of gpu for academic use.
Crypto is an example.
jxf
on Feb 4, 2023
|
next
[–]
This creates indirection costs and counterparty risks that don't appear in the original solution.
ec109685
on Feb 4, 2023
|
parent
|
next
[–]
There is also indirection cost by taking something that is optimized to run on GPU’s within the data center and distributing that to individual PCs.
slim
on Feb 4, 2023
|
prev
[–]
this would be very wasteful
ec109685
on Feb 4, 2023
|
parent
[–]
So is trying to distribute training across nodes compared to what can be done inside a data center.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
Crypto is an example.