Skip to content

feat(sdk): isolate static tensor arena configuration to protect SDK from vendor modifications#13

Open
lalalaorz wants to merge 1 commit into
edgeimpulse:masterfrom
lalalaorz:static_tensor_arena
Open

feat(sdk): isolate static tensor arena configuration to protect SDK from vendor modifications#13
lalalaorz wants to merge 1 commit into
edgeimpulse:masterfrom
lalalaorz:static_tensor_arena

Conversation

@lalalaorz

Copy link
Copy Markdown

Description

  • Introduce ei_get_static_tensor_arena as an extensible weak interface, allowing developers to seamlessly control arena memory attributes (alignment, linker sections, size) from the application layer.
  • Eliminate the need for users to modify SDK source code when porting to specialized hardware platforms (e.g., configuring non-cacheable SRAM regions).
  • Adapt the TFLite Micro interpreter initialization to properly ingest the new ei_unique_ptr_t smart pointer interface under static allocation mode.

…rom vendor modifications

- Introduce `ei_get_static_tensor_arena` as an extensible weak interface,
  allowing developers to seamlessly control arena memory attributes (alignment,
  linker sections, size) from the application layer.
- Eliminate the need for users to modify SDK source code when porting to
  specialized hardware platforms (e.g., configuring non-cacheable SRAM regions).
- Adapt the TFLite Micro interpreter initialization to properly ingest the
  new `ei_unique_ptr_t` smart pointer interface under static allocation mode.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant