Hi! I have been advocating the use of a threadpool as implemented in the Intel TBB as opposed to using our current C++11 threading model which creates new threads everytime one dispatches asynchronous work. I finally have a good way to benchmark this and to observe the speed difference. I am using…

Benefits of parallelization with a threadpool of the Intel TBB

andrjohns July 15, 2019, 12:17am 5

I can also test on my system (ryzen 2700x) if you put up an example to run

Threading and mpi and tbb and gpu in cmdstan

Topic		Replies	Views
Threading and mpi and tbb and gpu in cmdstan CmdStan	4	772	October 30, 2019
What would happen if no. of shards or partial sums is greater than threads per chain? CmdStan	4	359	July 29, 2020
Optimal num_stan_threads when using multiple chains General performance	5	1740	May 30, 2019
Reduce_sum() using only one thread Interfaces	2	399	June 10, 2020
Comparing threading and MPI General	10	2201	March 5, 2019