DIMM-Link: Enabling Efficient Inter-DIMM Communication for Near-Memory Processing
Zhe Zhou, Cong Li, Fan Yang, Guangyu Sun
Abstract
DIMM-based near-memory processing architectures (DIMM-NMP) have received growing interest from both academia and industry. They have the advantages of large memory capacity, low manufacturing cost, high flexibility, compatible form factor, etc. However, inter-DIMM communication (IDC) has become a critical obstacle for generic DIMM-NMP architectures because it involves costly forwarding transactions through the host CPU. Recent research has demonstrated that, for many applications, the overhead induced by IDC may even offset the performance and energy benefits of near-memory processing.To tackle this problem, we propose DIMM-Link, which enables high-performance IDC in DIMM-NMP architectures and supports seamless integration with existing host memory systems. It adopts bidirectional external data links to connect DIMMs, via which point-to-point communication and inter-DIMM broadcast are efficiently supported in a packet-routing way. We present the full-stack design of DIMM-Link, including the hardware architecture, interconnect protocol, system organization, routing mechanisms, optimization strategies, etc. Comprehensive experiments on typical data-intensive tasks demonstrate that the DIMM-Link-equipped NMP system can achieve a 5.93× average speedup over the 16-core CPU baseline. Compared to other IDC methods, DIMM-Link outperforms MCN, AIM, and ABC-DIMM by 2.42×, 1.87×, and 1.77×, respectively. More importantly, DIMM-Link fully considers the implementation feasibility and system integration constraints, which are critical for designing NMP architectures based on modern DDR4/DDR5 DIMMs.
Ask about this paper
Ask your agent about it.
Lune has read the top-tier papers around this one, so every answer names the papers it rests on.
Your agent calls
Lunesearch_papers
Free to start. No credit card required.
Terminal
Install the CLIlune papers get 8359189b-7231-48ca-a971-f252785da4baCited by top-tier papers13
- Accelerating Retrieval-Augmented GenerationDerrick Quinn, Mohammad Nouri, Neel Patel, John Salihu et al.ASPLOS 2025 · 37 citations
- PIM-DL: Expanding the Applicability of Commodity DRAM-PIMs for Deep Learning via Algorithm-System Co-OptimizationCong Li, Zhe Zhou, Yang Wang, Fan Yang et al.ASPLOS 2024 · 27 citations
- NDPBridge: Enabling Cross-Bank Coordination in Near-DRAM-Bank Processing ArchitecturesBoyu Tian, Yiwei Li, Li Jiang, Shuangyu Cai et al.ISCA 2024 · 27 citations
- Low-Overhead General-Purpose Near-Data Processing in CXL Memory ExpandersHyungkyu Ham, Jeongmin Hong, Geonwoo Park, Yunseon Shin et al.MICRO 2024 · 26 citations
- Make LLM Inference Affordable to Everyone: Augmenting GPU Memory with NDP-DIMMLian Liu, Shixin Zhao, Bing Li, Haimeng Ren et al.HPCA 2025 · 15 citations
Related papers
- ABC-DIMM: Alleviating the Bottleneck of Communication in DIMM-based Near-Memory Processing with Inter-DIMM BroadcastWeiyi Sun, Zhaoshi Li, Shouyi Yin, Shaojun Wei et al.ISCA 2021 · 42 citations
- AsyncDIMM: Achieving Asynchronous Execution in DIMM-Based Near-Memory ProcessingLiyan Chen, Dongxu Lyu, Jianfei Jiang, Qin Wang et al.HPCA 2025 · 7 citations
- HAIL-DIMM: Host Access Interleaved with Near-Data Processing on DIMM-based Memory SystemMinkyu Lee, Sang-Seol Lee, Kyungho Kim, Eunchong Lee et al.DAC 2024 · 2 citations
- Design and Analysis of a Processing-in-DIMM Join Algorithm: A Case Study with UPMEM DIMMsChaemin Lim, Suhyun Lee, Jinwoo Choi, Jounghoo Lee et al.SIGMOD 2023 · 50 citations
- PID-Comm: A Fast and Flexible Collective Communication Framework for Commodity Processing-in-DIMM DevicesSi Ung Noh, Junguk Hong, Chaemin Lim, Seongyeon Park et al.ISCA 2024 · 12 citations
