Search results

If multiple Einsums are run at once, don't double count leak power for shared components, and don't directly sum leak energy

Status: Open.

#32 In Accelergy-Project/accelforge;

· tanner-andrulis opened on Feb 21, 2026

analyze_compute in symbolic reuse analysis still goes through all tensors instead of the current tensor analyzed for the pmapping.

Status: Open.

#30 In Accelergy-Project/accelforge;

· gilbertmike opened on Feb 17, 2026

Implement "operations" key in workload Einsums

Status: Open.

#29 In Accelergy-Project/accelforge;

· tanner-andrulis opened on Feb 10, 2026

Making Unfused Loops Fused

Status: Open.

#27 In Accelergy-Project/accelforge;

· tanner-andrulis opened on Jan 29, 2026

Allow backing storage be above in the mapping anything that is below it in the memory hierarchy

Status: Open.

#24 In Accelergy-Project/accelforge;

· tanner-andrulis opened on Jan 13, 2026

Optimization for conv workload mapping

Status: Open.

#21 In Accelergy-Project/accelforge;

· tanner-andrulis opened on Jan 7, 2026

Some constraint processing should be moved out of arch.py and _workload_isl needs refactor

Status: Open.

#18 In Accelergy-Project/accelforge;

· tanner-andrulis opened on Jan 5, 2026

More citations needed in citation.rst

Status: Open.

#17 In Accelergy-Project/accelforge;

· tanner-andrulis opened on Jan 5, 2026