Skip to content

HIP vs OpenCL #186

Description

@luminousmining

Maintaining both OpenCL and CUDA is starting to become difficult.
A study should be done on deprecating OpenCL in this miner and using HIP exclusively for AMD cards.
This would allow using the same CUDA kernels for the same devices.

This opens the debate for Intel GPUs — OpenCL or SYCL?

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    amddocumentationImprovements or additions to documentationhelp wantedExtra attention is neededintel

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions