Full Program

Day 1 — September 28, 2026

13:30 – 13:40	Opening Remarks
13:40 – 14:40	Session 1: Storage System Software
14:40 – 15:00	Coffee Break
15:00 – 16:00	Session 2: Memory and Storage
16:00 – 16:20	Coffee Break
16:20 – 17:30	Session 3: Solid-State Drives and Poster Talks

Day 2 — September 29, 2026

9:00 – 10:00	Keynote
10:00 – 10:30	Coffee Break
10:30 – 11:50	Session 4: New Technology
12:00 – 13:30	Lunch
13:30 – 15:10	Session 5: Disaggregated and Distributed
15:10 – 15:30	Coffee Break
15:30 – 17:10	Session 6: Storage in AI

Sessions

Session 1: Storage System Software

• To Migrate or Not to Migrate: KOxI, an Evidence-Gated Pipeline for Rust-for-Linux Storage Drivers
• Manycore Scalable Logging for Log-Structured Filesystem
• Print-Time Audit-Log Integrity Below the Operating System

Session 2: Memory and Storage

• PMEM Throughput Aggregation with Non-volatile Page Table Tree
• User Space Buffer Cache with the Bells and Whistles of a Kernel Page Cache
• CopyCat: Harvesting the Frequency Tax of Bulk Memory Copy

Session 3: Solid-State Drives and Poster Talks

• The Host Is Not Idle: Computational SSDs for Host-Free Checkpointing in Multimodal LLM Training
• GraceSSD: Per-File Reliability Hints for Error-Tolerant Machine Learning Storage
• io.stream: cgroup-Granular FDP for Unmodified Workloads

• Exploring High-Bandwidth Flash in GPU Memory Systems with HBFSim
• Towards 100 Million IOPS for GPU-initiated I/O

Session 4: New Technology

• I/O-Cool: Using I/Os for Temperature Regulation When Training on the Edge
• The Storage Wall in LEO Compute Clouds
• Towards A System Model for DNA-based Storage
• CXL Memory for LLM Inference Staging via an On-Device DMA Controller

Session 5: Disaggregated and Distributed

• RLSM: A Disaggregated LSM-Based Key-Value Store with Resource-Aware Compaction
• Where Should Compaction Run? Compaction Worker Selection on Disaggregated LSM-KVS
• MQSim 2.0: A Framework for Realistic Studies of AI-Era SSDs and Disaggregated Storage
• ZWAN: Leveraging Zone Random Write Area (ZRWA) for Avoiding WAL Tax in LSM-tree
• Asclepius: Duplication and Heterogeneity-Aware Data Repair in Erasure-Coded Storage Systems

Session 6: Storage in AI

• Bridging CPU and GPU I/O with a Device-Resident File System
• PLINK: A GPU-Initiated I/O Platform Exploiting NVMe Parallelism
• Token Write Amplification in LLM-Mediated State Stores
• Disaggregated LLM KV-Cache Storage Requires Coordination-Free Consistency Abstractions
• LLM KV-Cache: To Restore or To Recompute, That Is the Question