Mxnet R Installation for GPU (Windows) not working

I am trying to install the GPU version in R and folow the official instructions:
cran <- getOption(“repos”)
cran[“dmlc”] <- “
options(repos = cran)

is not avaliable.

the CPU version works without problems

Am I the only one with this problem?

Hi GeorgeKaplan,
Could you please tell me how did you try to install GPU version. There are some pre requisite like installing Nvidia drivers and cuda toolkit.
Please take a look at " Installing MXNet on a Computer with a GPU Processor" in

Please let me know if it works.

Hi ankkhedia!
I was exactly following the steps provided in the link you posted

### Installing MXNet on a Computer with a GPU Processor
Nvidea drivers installed, Cuda installed, VS installed, CuDNN installed.

However I dont think that is the problem because when I follow the steps:

#### Installing MXNet with the Prebuilt Binary Package(GPU)
and use the code in R:

cran <- getOption(“repos”)
_ cran[“dmlc”] <- “”_
_ options(repos = cran)_
_ install.packages(“mxnet”)_

R Studio simply tells me that the file is not avaliable:

trying URL ‘xttps://’
Warning in install.packages :
cannot open URL ‘xttps://’: HTTP status was ‘404 Not Found’

I had to change https to xttps because new users are only allowed to put 2 links in one post
Thank you for your time

Can you please post the result from sessionInfo() ?

R version 3.4.0 (2017-04-21)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows >= 8 x64 (build 9200)

Matrix products: default

[1] LC_COLLATE=German_Austria.1252 LC_CTYPE=German_Austria.1252 LC_MONETARY=German_Austria.1252 LC_NUMERIC=C
[5] LC_TIME=German_Austria.1252

attached base packages:
[1] stats graphics grDevices utils datasets methods base

loaded via a namespace (and not attached):
[1] compiler_3.4.0 tools_3.4.0 yaml_2.2.0

So far the GPU versions are compiled with R 3.5.0. One fast solution is to update your R version to 3.5 and retry the installation.

Thanks a lot.
So easy and so difficult.

…maybe this could be added to the installation page.

1 Like