mxnet for GPU with CUDA 11

[manybodycpa]

This is not strictly related to insightface but is there a GPU-enabled version of mxnet available for use with CUDA 11?

[xxxpsyduck]

wait for mxnet 2.0

[Light--]

Isn't the nvidia official said that cuda 11 is compatible with mxnet 1.6.0 ?

Contents of the MXNet container
This container image contains the complete source of the version of NVIDIA Optimized Deep Learning Framework, which is based upon Apache **MXNet version 1.6.0.** It is pre-built and installed to the Python path.

The container also includes the following:
Ubuntu 18.04 including Python 3.6
**NVIDIA CUDA 11.0.167 including cuBLAS 11.1.0**
NVIDIA cuDNN 8.0.1
NVIDIA NCCL 2.7.5 (optimized for NVLink™ )

the link is here:
https://docs.nvidia.com/deeplearning/frameworks/mxnet-release-notes/rel_20-06.html

-----------------------------update 11/16/20---------------
actually i tried this: https://docs.nvidia.com/deeplearning/frameworks/mxnet-release-notes/rel_20-10.html#rel_20-10 and core dumped.... 😞

[evo11x]

Sam problem here, is there a way to use CUDA 10 on RTX 3060 ?

[matheus695p]

same problem here, im trying to use gluonts on rtx 3080 :/

[nttstar]

@szha

[szha]

mxnet 1.8.0.post includes support for both cu110 and cu112

[roachsinai]

mxnet 1.8.0.post includes support for both cu110 and cu112

yes, but no support for windows.

[dennis-gdf]

If you use windows 11, WSL could be a solution. WSL now uses host GPU device directly via MS version linux kernel.