Solidigm CSAL Solution Brings Advanced IO Shaping, Caching and Data Placement Into NVIDIA DPU DOCA S
Talkat KubeCon + CloudNativeCon China 2025 · on Wednesday, 2025-06-11· at 13:45· Level 19 | Crystal Court II·
summary
Describes CSAL (Cloud Storage Acceleration Layer), an open-source user‑mode FTL, cache and IO tracing component inside SPDK, and how it’s integrated into NVIDIA DPU DOCA storage services alongside BeeGFS. Covers using DPU DRAM as a CSAL write buffer for low-latency, consistent storage, support for high‑density QLC storage for AI datacenters, and experimental data from the joint Solidigm–NVIDIA–BeeGFS development.
AI-written summary · about
related talks
12mAdvancing Flash Storage
38mPOLARDB Meets Computational Storage: Efficiently Support Analytical Workloads in Cloud
15mHere, There and Everywhere: The Past, the Present and the Future of Local Storage in...
14mPD3: Prefetching Data with DPUs for Disaggregated Memory
17mDuhu: Shared Disaggregated Memory for Distributed Data Processing Frameworks
25mOpen Source Model Performance Optimization With SGLang
35mOpenSDS: Storage Challenges in a Cloud Native Era
19mDShuffle: DPU-Optimized Shuffle Framework for Large-scale Data Processing
