Skip to content

help: gcc + openacc + nvidia dgup #43

Open
@finjulhich

Description

This is more a help question rather than a issue

I have a Linux Fedora 38 system with a very old Nvidia GF119M [NVS 4200M] (rev a1)
The nvidia driver is installed and these packages are installed

gcc-13.2.1-4.fc38.x86_64
gcc-offload-nvptx-13.2.1-4.fc38.x86_64
libgomp-offload-nvptx-13.2.1-4.fc38.x86_64

xorg-x11-drv-nvidia-390xx-kmodsrc-390.157-2.fc38.x86_64
xorg-x11-drv-nvidia-390xx-libs-390.157-2.fc38.x86_64
xorg-x11-drv-nvidia-390xx-390.157-2.fc38.x86_64
xorg-x11-drv-nvidia-390xx-cuda-libs-390.157-2.fc38.x86_64
xorg-x11-drv-nvidia-390xx-cuda-390.157-2.fc38.x86_64
akmod-nvidia-390xx-390.157-5.fc38.x86_64
kmod-nvidia-390xx-6.5.8-200.fc38.x86_64-390.157-5.fc38.x86_64
kmod-nvidia-390xx-6.5.9-200.fc38.x86_64-390.157-5.fc38.x86_64
nvidia-persistenced-535.113.01-1.fc38.x86_64

Am I as is then able to compile C++ code with OpenAcc directives and produce a binary that can run code on CPU + GPU?

or is there more to install

I found https://gcc.gnu.org/wiki/OpenACC quite hard to read

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions