Buffer Prospector: Discovering and Exploiting Untapped Buffer Resources in Many-Core DNN Accelerators
In large-scale DNN inference accelerators, the many-core architecture has emerged as a predominant design, with layer-pipeline (LP) mapping being a mainstream mapping approach. However, our experimental findings and theoretical justifications uncover a hardware-independent and prevalent flaw in empl...
Saved in:
| Published in: | 2025 62nd ACM/IEEE Design Automation Conference (DAC) pp. 1 - 7 |
|---|---|
| Main Authors: | , , , , , , |
| Format: | Conference Proceeding |
| Language: | English |
| Published: |
IEEE
22.06.2025
|
| Subjects: | |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In large-scale DNN inference accelerators, the many-core architecture has emerged as a predominant design, with layer-pipeline (LP) mapping being a mainstream mapping approach. However, our experimental findings and theoretical justifications uncover a hardware-independent and prevalent flaw in employing layer-pipeline mapping on many-core accelerators: a significant underutilization of buffer space across numerous cores, indicating substantial potential for optimization. Building on this discovery, we develop a universal and efficient buffer allocation strategy, BufferProspector, which includes a Buffer Requirement Calculator and Buffer Allocator, to capitalize on these unused buffers, addressing the timing mismatch challenge inherent in LP mapping. Compared to the state-of-the-art (SOTA) open-source LP mapping framework Tangram, BufferProspector averages a simultaneous increase in energy efficiency and performance by 1.44× and 2.26×, respectively. Moreover, we conduct some case studies on architecture and mapping. BufferProspector will be open-sourced. |
|---|---|
| DOI: | 10.1109/DAC63849.2025.11132440 |