- #1
Wesleytf
- 31
- 0
hello all! I just got a new computer with an Nvidia card, and am now able to do some GPU parallel processing inside mathematica using CUDA.
My main interest is in taking tons of moderate accuracy (3-4 digits) numerical integrals. I've been using QMC in MMA and that's been working well, but I'd like to try it on the GPU to hopefully expedite the process. This would be trivial if they'd included that as one of the built in functions, but alas, they did not.
Has anyone on here attempted doing something like this? After reviewing the documentation, it looks like something I wouldn't be able to implement without some help.
My main interest is in taking tons of moderate accuracy (3-4 digits) numerical integrals. I've been using QMC in MMA and that's been working well, but I'd like to try it on the GPU to hopefully expedite the process. This would be trivial if they'd included that as one of the built in functions, but alas, they did not.
Has anyone on here attempted doing something like this? After reviewing the documentation, it looks like something I wouldn't be able to implement without some help.