Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2512.09304 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917135935078400 |
|---|---|
| author | Ma, Siyuan Hu, Jiajun Ryoo, Jeeho Arora, Aman John, Lizy Kurian |
| author_facet | Ma, Siyuan Hu, Jiajun Ryoo, Jeeho Arora, Aman John, Lizy Kurian |
| contents | In-DRAM Processing-In-Memory (DRAM-PIM) has emerged as a promising approach to accelerate memory-intensive workloads by mitigating data transfer overhead between DRAM and the host processor. Bit-serial DRAM-PIM architectures, further enhance efficiency by supporting runtime variable data precision, which is critical for emerging workloads, such as large language model (LLM) inference. However, existing works still have major limitations: lack of data reuse, significant amounts of redundant data transfer, and insufficient support for workload mapping. To address these issues, we propose RACAM, the first in-DRAM bit-serial architecture which uses dedicated locality buffers, bit-serial PEs, popcount reduction units and broadcast units to enable data reuse and alleviate redundant data transfers. Furthermore, a workload mapping mechanism is proposed to fully explore the massive parallelism of DRAM architecture and identify the best mapping scheme of a given workload. We evaluate RACAM against GPUs and the state-of-the-art, in-DRAM PIM system, Proteus, across end-to-end LLM inferences. RACAM achieves 9x to 102x speedup over GPUs and 233x higher performance per mm2 compared to Proteus in case of GPT3. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_09304 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RACAM: Enhancing DRAM with Reuse-Aware Computation and Automated Mapping for ML Inference Ma, Siyuan Hu, Jiajun Ryoo, Jeeho Arora, Aman John, Lizy Kurian Hardware Architecture In-DRAM Processing-In-Memory (DRAM-PIM) has emerged as a promising approach to accelerate memory-intensive workloads by mitigating data transfer overhead between DRAM and the host processor. Bit-serial DRAM-PIM architectures, further enhance efficiency by supporting runtime variable data precision, which is critical for emerging workloads, such as large language model (LLM) inference. However, existing works still have major limitations: lack of data reuse, significant amounts of redundant data transfer, and insufficient support for workload mapping. To address these issues, we propose RACAM, the first in-DRAM bit-serial architecture which uses dedicated locality buffers, bit-serial PEs, popcount reduction units and broadcast units to enable data reuse and alleviate redundant data transfers. Furthermore, a workload mapping mechanism is proposed to fully explore the massive parallelism of DRAM architecture and identify the best mapping scheme of a given workload. We evaluate RACAM against GPUs and the state-of-the-art, in-DRAM PIM system, Proteus, across end-to-end LLM inferences. RACAM achieves 9x to 102x speedup over GPUs and 233x higher performance per mm2 compared to Proteus in case of GPT3. |
| title | RACAM: Enhancing DRAM with Reuse-Aware Computation and Automated Mapping for ML Inference |
| topic | Hardware Architecture |
| url | https://arxiv.org/abs/2512.09304 |