Skip to content

memref-merge optimization #269

Closed
Closed
@ciyongch

Description

@ciyongch
  1. A new pass is required to reduce the total memory footprint and improve the cache locality.
  2. The static memory planner is required to get the total memory size and the map to each tensor during the compilation stage.
  3. The pass is extensible to support different event based ticker and planner.

Metadata

Metadata

Assignees

Labels

Type

No type

Projects

No projects

Relationships

None yet

Development

No branches or pull requests

Issue actions