This topic arose in advising a user seeking to accelerate their sampling using cloud services. At first I suggested they look into reduce_sum(), but then remembered that the GPU crew implemented accelerators for the likelihood computation for GLMs. Do we have any intuitions or results yet on when it…

Any intuitions/results on when GPU>reduce_sum?

saudiwin December 18, 2020, 6:57pm 5

What happens if you combine reduce_sum with gpu support? Will it offload matrix calculations to the GPU within threads? Or does that just get turned off?

Categorical model with large dataset - large/divergent ELBO & possible combined reduce_sum()/GPU support

Topic		Replies	Views
Reduce-sum parallelization General techniques , paralellization	18	160	August 1, 2024
Reduce_sum performance Modeling	5	864	May 22, 2020
Reduce_sum results in much slower run times, even for large datasets Algorithms paralellization	6	1406	March 17, 2022
Understanding reduce_sum efficiency Modeling	10	844	March 22, 2021
Reduce_sum performance Modeling performance , paralellization	15	1342	September 28, 2020

Any intuitions/results on when GPU>reduce_sum?

Related topics