GNStor is a GPU-native AFA system that enables direct GPU access to remote AFA without CPU intervention.
GNStor is novel in its ability to enable direct GPU access to remote AFA without CPU intervention, which improves I/O performance and reduces latency.
Before reading this…
Applications
To understand this paper, make sure you know these concepts first:
GPU has become the leading computing device for a wide range of data-intensive applications, which tightly collaborates with remote all-flash array (AFA) to accommodate ever-expanding datasets, facilitate multi-client data sharing, and guarantee fault tolerance. Although GPU is the center of computation, all I/O processes in existing GPU-AFA systems are still CPU-centric. CPU orchestrates remote I/O requests and executes a centralized AFA engine to take charge of AFA-level functionalities (e.g., access control and metadata persistence). This design disparity suffers from substantial CPU-GPU interaction overhead and I/O traffic amplification, compromising end-to-end I/O performance. In this work, we present \emph{GNStor}, a GPU-native AFA system that enables GPU to directly access remote AFA without CPU intervention in the I/O path, thereby fully exploiting the performance of AFA. Specifically, GNStor first proposes a GPU-centric NVMe over RDMA (NoR) software stack (named \emph{GNoR}), paving a fast path for GPUs to directly initiate NoR I/O requests to SSDs within remote AFA. GNoR employs an atomic-operation-based I/O orchestration design and follows the single-instruction-multiple-thread (SIMT) execution model of GPU, fully exploiting the massive parallelism of GPU architectures. To facilitate essential AFA functionalities in a CPU-bypass I/O path, GNStor further designs \emph{deEngine}, a decentralized AFA engine that seamlessly decomposes and integrates AFA-level tasks into each SSD firmware, thereby achieving efficient AFA access at low cost. Evaluation results show that GNStor achieves 3.2$\times$ higher I/O throughput and reduces application execution time by 31.1\%, compared to state-of-the-art AFA systems.
NELSSA: A GPU-PNM Heterogeneous System for Mixed-Length LLM Serving via Length-based Request Placeme…
This paper presents NELSSA, a serving system that integrates GPUs with Processin…
GPU Offload in Rust: Portable, Safe, and Fast
This paper presents a rustc-based GPU compilation framework for Rust, ensuring m…
NPU Offloading of a Frozen Visual Encoder for Robot Policy Training
This paper investigates the energy savings and performance impact of offloading…
DiffPower: GPU-Accelerated Differentiable Switching Power Analysis and Optimization
The paper presents DiffPower, a GPU-accelerated framework for differentiable pow…
HetRoute Heterogeneous and Cost-aware Collaborative Routing Framework for Distributed Edge MoE Infer…
This paper proposes HetRoute, a framework for optimizing Mixture-of-Experts mode…
Exploring High-Bandwidth Flash for Modern LLM Inference: Opportunities and Challenges
This paper investigates the benefits and challenges of using high-bandwidth flas…
What Actually Serializes GPU LZ77 Decode: Three Decoders, Three Mechanisms, and an Encode-Time Lever…
This paper challenges assumptions about the sequential part of GPU LZ77 decode a…
TokaGLINT: A Scalable GPU-Tailored Implicit Solver for Full 3D Tokamak Electromagnetic Simulations
This paper introduces TokaGLINT, a GPU-accelerated solver for electromagnetic fi…