Let us say a MFLOP has an energetic cost of C. Then M * MFLOP * C is the training cost, and sqrt(M) * MFLOP * C is the query cost. As pointed out above, if the speed of the system serving the query is in the GFLOP/s range and it spends a few seconds on it, the square-root argument is not too outlandish, given that it took similar system days to train.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| AI models are too costly | 78 | 2515 | December 27, 2023 | |
| I Have a dream! a Green dream -- Does Julia save energy? | 21 | 3141 | December 18, 2021 | |
| How much faster is GPU compare to CPU | 16 | 27530 | November 24, 2018 | |
| How to measure energy consumption of computations? | 7 | 1097 | May 6, 2021 | |
| Two questions on Flux | 23 | 4940 | October 2, 2020 |