[Rd] Linking to Intel's MKL on Windows
@vr@h@m@@d|er @end|ng |rom gm@||@com
Sun Oct 2 01:52:39 CEST 2022
Also, you can build Rblas against OpenBLAS even on Windows which will go far in speeding up matrix calculations.
Sent from my iPhone
> On Oct 1, 2022, at 12:49 PM, Ben Bolker <bbolker using gmail.com> wrote:
> Maybe you can find out more about Microsoft's development/release process for MRO and why they're still on 4.0.2 (from June 2020)? I followed the "user forum" link on their web page, but it appears to be a generic Windows forum ...
> I might tweet at @revodavid (David Smith) to see if there's any more information available about the MRO release schedule ...
> good luck,
> Ben Bolker
>> On 2022-10-01 12:00 p.m., Viechtbauer, Wolfgang (NP) wrote:
>> Hi Christine,
>> MKL is a closed-source commercial product (yes, one can get it for free, but it is not libre/open-source software).
>>> -----Original Message-----
>>> From: R-devel [mailto:r-devel-bounces using r-project.org] On Behalf Of Christine
>>> Stawitz - NOAA Federal via R-devel
>>> Sent: Friday, 30 September, 2022 18:46
>>> To: r-devel using r-project.org
>>> Subject: [Rd] Linking to Intel's MKL on Windows
>>> Recently I became aware that Microsoft R Open provides accelerated matrix
>>> algebra computations through Intel's Math Kernel Libraries. However, the
>>> version of R shipped with the Microsoft R Open is too out of date to be
>>> able to use concurrently with other dependencies while developing our
>>> package. This thread suggests a way to get the updated matrix libraries
>>> with a more recent version of R, however it is unlikely to be approved by
>>> our IT admin since those of us in government agencies aren't typically
>>> given admin privileges: Linking Intel's Math Kernel Library (MKL) to R on
>>> Windows - Stack Overflow
>>> Is there a reason why CRAN doesn't provide a version of R with the updated
>>> libraries such that developers don't have to recompile R or copy .dlls
>>> around as described above? It would help those of us running software with
>>> slow-running matrix calculations in R.
>>> Christine C. Stawitz, PhD. (pronouns: she/her)
>>> National Stock Assessment Program Modeling Team
>>> NOAA Fisheries Office of Science and Technology | U.S. Department of
>>> Mobile: 206-617-2060
>> R-devel using r-project.org mailing list
> Dr. Benjamin Bolker
> Professor, Mathematics & Statistics and Biology, McMaster University
> Director, School of Computational Science and Engineering
> (Acting) Graduate chair, Mathematics & Statistics
> > E-mail is sent at my convenience; I don't expect replies outside of working hours.
> R-devel using r-project.org mailing list
More information about the R-devel