From the 1 of 16 linked papers with an AI index.
16 papers
The Fabric Is the Cluster Driver: Cross-Layer eBPF Policies for GPU-CXL Fabrics
Yiwei Yang, Andi Quinn
The paper introduces fabric_ext, an eBPF‑based middleware that lets developers write cross‑layer policies spanning GPUs, DPUs/NICs, and CXL switches, using a semantic movement grap…
Concordia: JIT-Compiled Persistent-Kernel Checkpointing for Fault-Tolerant LLM Inference
Yuhang Gan, Yiwei Yang, Yuyi Li +6
Long-running LLM agents keep valuable state resident on GPUs: KV caches, request schedulers, communication state, and sometimes online adapters. Losing this state after a GPU or co…
CXLMemUring: A Hardware Software Co-design Paradigm for Asynchronous and Flexible Parallel CXL Memory Pool Access
Yiwei Yang, Yusheng Zheng, Kexin Chu +3
CXL-attached memory lets servers add more memory while keeping the standard load/store programming model. The main drawback is latency. CXL memory accesses are too slow for normal…
GPUOS: A GPU Operating System Primitive for Transparent Operation Fusion
Yiwei Yang, Xiangyu Gao, Yuan Zhou +3
Modern deep learning workloads often consist of many small tensor operations, especially in inference, attention, and micro-batched training. In these settings, kernel launch overh…
WIO: Upload-Enabled Computational Storage on CXL SSDs
Yiwei Yang, Yanpeng Hu, Yusheng Zheng +4
The widening gap between processor speed and storage latency has made data movement a dominant bottleneck in modern systems. Two lines of storage-layer innovation attempted to clos…
DAXFS: A Lock-Free Shared Filesystem for CXL Disaggregated Memory
Cong Wang, Yiwei Yang, Yusheng Zheng
CXL (Compute Express Link) enables multiple hosts to share byte-addressable memory with hardware cache coherence, but no existing filesystem exploits this for lock-free multi-host…