Skip to content

OpenCL: large graph support through RAM->VRAM streaming #211

Description

@EgorOrachyov

Requirements:

  • Async streaming of chunks of graph from RAM into VRAM
  • Dynamic chunk size adaptation
  • Partial operations evaluation support

Metadata

Metadata

Labels

enhancementNew feature or requesthelp wantedExtra attention is neededtopic:apiLibrary API relatedtopic:gpuOpenCL (Compute) relatedtopic:perfPerformance optimziations

Type

No type

Projects

No projects

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions