Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Compile EinExprs to native code or IR #13

Open
mofeing opened this issue May 11, 2023 · 0 comments
Open

Compile EinExprs to native code or IR #13

mofeing opened this issue May 11, 2023 · 0 comments
Assignees
Labels
enhancement New feature or request performance Makes the code go "brrrr"

Comments

@mofeing
Copy link
Member

mofeing commented May 11, 2023

EinExpr can be seen as symbolic expressions that represent linear operations expressible using the Einstein summation notation.

The user can walk through the tree and execute the appropriate methods by identifying the array types with almost no overhead.

But what if we could compile directly these symbolic trees into native code? Could we gain something?

Things to check:

  • Link Backward Automatic Differentiation with compilation
  • Link it with batched contraction
  • Multiple Level Intermediate Representation (MLIR)
@mofeing mofeing added enhancement New feature or request performance Makes the code go "brrrr" labels May 11, 2023
@mofeing mofeing added this to the 0.2 milestone May 11, 2023
@mofeing mofeing self-assigned this May 11, 2023
@mofeing mofeing changed the title Compile EinExprs Compile EinExprs to native code May 11, 2023
@mofeing mofeing changed the title Compile EinExprs to native code Compile EinExprs to native code or IR May 11, 2023
@mofeing mofeing removed this from the 0.2 milestone Jul 22, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request performance Makes the code go "brrrr"
Projects
None yet
Development

No branches or pull requests

1 participant