Simon, That is very helpful. Your advice about the memory management is what I do in bigCor to allow me to find covariances that would otherwise blow up. It takes advantage of multiple cores that we have on Macs and allows me to solve problems that I could not handle directly. I had the hope that some magic would help me in using the GPU to speed things up and I had just missed finding the approrpriate package. From your note, I have not.
Many of us R for Mac users are forever in your debt for your support of R for Macs. Bill > On Feb 21, 2023, at 4:31 PM, Simon Urbanek <simon.urba...@r-project.org> > wrote: > > Bill, > > to answer the general part of you question, there are quite a few packages > that allow you to use your GPU. At the low level you can use OpenCL which > allows you to write code directly on the GPU and turn it into R functions. At > a higher level there are frameworks like Torch and Tensorflow that support > GPUs and there are packages like keras or rTorch that provide interfaces to > that, e.g. for machine learning models. All those support Apple silicon GPUs. > > However, you mentioned bigCor which is orthogonal to that. There are papers > (such as https://doi.org/10.1134/S1995080219050068 ) that describe some > approaches to using GPUs to compute covariance matrices, but in general GPUs > don't really give you a large benefit on their own since it is a lot more > about memory management and using smart algorithms/data structures to reduce > the necessary computation. You may have more luck asking in HPC forums about > that since is not really Mac-specifc (and the answer may depend on what you > are actually trying to do so may need more details). > > Cheers, > Simon > > >> On 22/02/2023, at 6:12 AM, William R Revelle <reve...@northwestern.edu> >> wrote: >> >> Dear R-Mac users. >> >> In trying to speed up a large correlation problem (600K subjects, 6k >> variables,) which I can do using my bigCor function, I decided it was time >> to learn how to use GPU on my Mac book with its M1 Max gpu. >> >> Having spent a day searching the web and trying various approaches, I give >> up. >> >> Are there any packages I can use to do calculations on the GPU part of my >> Mac using R? >> >> Thanks. >> >> Bill >> >> William Revelle personality-project.org/revelle.html >> Professor personality-project.org >> Department of Psychology www.wcas.northwestern.edu/psych/ >> Northwestern University www.northwestern.edu/ >> Use R for psychology personality-project.org/r >> It is 90 seconds to midnight www.thebulletin.org >> >> _______________________________________________ >> R-SIG-Mac mailing list >> R-SIG-Mac@r-project.org >> https://stat.ethz.ch/mailman/listinfo/r-sig-mac >> > > _______________________________________________ > R-SIG-Mac mailing list > R-SIG-Mac@r-project.org > https://stat.ethz.ch/mailman/listinfo/r-sig-mac William Revelle personality-project.org/revelle.html Professor personality-project.org Department of Psychology www.wcas.northwestern.edu/psych/ Northwestern University www.northwestern.edu/ Use R for psychology personality-project.org/r It is 90 seconds to midnight www.thebulletin.org _______________________________________________ R-SIG-Mac mailing list R-SIG-Mac@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-mac
