LightDSA: Enabling Efficient DSA Through Hardware-Aware Transparent Optimization
Yuansen Wang, Teng Ma, Yuanhui Luo, Dongbiao He, Zheng Liu, Yunpeng Chai
2026年份
摘要
Data streaming operations consume a significant portion of CPU resources in data centers. The Data Streaming Accelerator (DSA), integrated into modern Intel CPUs in datacenter, offers promising acceleration for these operations with user-friendly features. However, previous studies have overlooked DSA's internal mechanisms and the performance implications of these features, leaving key performance issues unresolved in real-world usage.
问问这篇 Paper
问问你的智能体。
Lune 读过与它相关的顶会 Paper,每个回答都会注明依据哪几篇。
相关 Paper
- A Quantitative Analysis and Guidelines of Data Streaming Accelerator in Modern Intel Xeon Scalable ProcessorsReese Kuper, Ipoom Jeong, Yifan Yuan, Ren Wang 等ASPLOS 2024 · 被引用 24 次
- DarkStream: Exploiting Internal Throughput Contention in Data Streaming Accelerator for Timing AttacksHyosang Kim, Ki-Dong Kang, Gyeongseo Park, Sungju Kim 等ISCA 2026
- Para-ksm: Parallelized Memory Deduplication with Data Streaming AcceleratorHouxiang Ji, Minho Kim, Seonmu Oh, Daehoon Kim 等USENIX ATC 2025 · 被引用 7 次
- DSAssassin: Cross-VM Side-Channel Attacks by Exploiting Intel Data Streaming AcceleratorBen Chen, Kunlin Li, Shuwen Deng, Dongsheng Wang 等HPCA 2026 · 被引用 1 次
- Data Motion Acceleration: Chaining Cross-Domain Multi AcceleratorsShu-Ting Wang, Hanyang Xu, Amin Mamandipoor, Rohan Mahapatra 等HPCA 2024 · 被引用 10 次
