← AI compute allocation
What podcasts say about AI compute allocation
Every statement, with the speaker, the exact quote and the moment it was said.
What experts have said about AI compute allocation
2 statements · 2 neutral
Total training-plus-inference cost is often minimized when component costs are equalized.
“the minimum tends to be where the costs are equalized”
Open the episode · Reiner Pope – The math behind how LLMs are trained and servedListen at 1:20:53
A rough compute allocation is one-third each for pretraining, RL, and inference.
“a good ballpark is 33% split between each of them”
Open the episode · Reiner Pope – The math behind how LLMs are trained and servedListen at 1:24:16
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.