Ascend-SpMM: Accelerating Sparse Matrix-Dense Matrix Multiplication on Ascend NPUs (Q7414962)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
No description defined
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Ascend-SpMM: Accelerating Sparse Matrix-Dense Matrix Multiplication on Ascend NPUs |
No description defined |
Statements
Ascend-SpMM: Accelerating Sparse Matrix-Dense Matrix Multiplication on Ascend NPUs (English)
0 references
1 November 2026
0 references
cs.DC
0 references
cs.DB
0 references
Jiahe Wang
0 references
Yingxia Shao
0 references
Shuhan Wang
0 references
Qingbo Ge
0 references
Hongzheng Li
0 references
Yucheng Zhang
0 references
We present a novel SpMM acceleration framework for modern NPUs featuring a decoupled architecture, introducing a memory-centric cost model and customized sub-kernels with hierarchy-aware pipelining and round-robin synchronization to achieve up to 20.55X speedup over the strongest baseline NPU implementation and competitive GPU performance.
0 references
Sparse Matrix Dense Matrix
0 references
NPU Acceleration
0 references
Memory-Centric Design
0 references
1 reference
