Abstract
In this paper, we propose an extension of the Processing-In-Memory (PIM) architecture's instruction set to optimize distance calculations. The existing PIM architecture, while efficient for neural network tasks, exhibits limitations in handling distance calculations due to frequent data transfers between memory banks. To address this, we introduced the custom instruction, reducing intermediate data storage needs and improving computation efficiency. Simulations using random matrices showed that our approach decreases cycle count by up to 44 %, with more significant performance gains for larger dataset. This extension demonstrates the potential of enhanced PIM architectures for efficient memory-bound operations.
| Original language | English |
|---|---|
| Title of host publication | 2025 International Conference on Electronics, Information, and Communication, ICEIC 2025 |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| ISBN (Electronic) | 9798331510756 |
| DOIs | |
| State | Published - 2025 |
| Event | 2025 International Conference on Electronics, Information, and Communication, ICEIC 2025 - Osaka, Japan Duration: 19 Jan 2025 → 22 Jan 2025 |
Publication series
| Name | 2025 International Conference on Electronics, Information, and Communication, ICEIC 2025 |
|---|
Conference
| Conference | 2025 International Conference on Electronics, Information, and Communication, ICEIC 2025 |
|---|---|
| Country/Territory | Japan |
| City | Osaka |
| Period | 19/01/25 → 22/01/25 |
Bibliographical note
Publisher Copyright:© 2025 IEEE.
Keywords
- PIM Simulator
- Processing In Memory
- vector search