Yijing Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Yijing Wang has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Jiexiong Xu
Zhejiang University, Hangzhou, China
3 shared publications - Wenzhi Chen
Zhejiang University, Hangzhou, China
3 shared publications - Yiquan Chen
Alibaba Cloud, Hangzhou, China
3 shared publications - Shushu Zhao
Alibaba Group, Hangzhou, China
2 shared publications - Wenhai Lin
Zhejiang University, Hangzhou, China
2 shared publications - Yiquan Lin
Zhejiang University, Hangzhou, China
2 shared publications
Latest Publications
- Open Access
- 19 May 2026
- Published by Association for Computing Machinery (ACM)
- p. 214-221
Ada-Store: An Adaptive Load-aware Hybrid Storage Architecture for Bursty Workloads
- 19 May 2026
- Published by Association for Computing Machinery (ACM)
- p. 214-221
Open AccessAda-Store: An Adaptive Load-aware Hybrid Storage Architecture for Bursty Workloads
- Open Access
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 750-765
OS2G: A High-Performance DPU Offloading Architecture for GPU-based Deep Learning with Object Storage
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 750-765
Open AccessOS2G: A High-Performance DPU Offloading Architecture for GPU-based Deep Learning with Object Storage
- Open Access
- 16 September 2024
- Published by Association for Computing Machinery (ACM)
- p. 14-24
Performance Characterization of SmartNIC NVMe-over-Fabrics Target Offloading
The NVMe-over-Fabrics (NVMe-oF) is gaining popularity in cloud data centers as a remote storage protocol for accessing NVMe storage devices across servers. With the rapid increase in throughput of the NVMe storage devices, the NVMe-oF stack consumes a significant amount of valuable CPU resources. To release these valuable computing power to other tasks, many smartNICs now support NVMe-oF Target offloading. However, this emerging NVMe-oF offloading scheme's performance has not been fully investigated. In this work, we conducted comprehensive evaluations on smartNIC NVMe-oF target offloading and non-offloading NVMe-oF with TCP and RDMA. Our results demonstrate that the target offloading can achieve comparable throughput compared to NVMe/RDMA in most synthetic and real-world application workloads while releasing up to 38.7% CPU resources. Additionally, target offloading benefits from host bypass and peer-to-peer communication between smartNIC and NVMe SSD, resulting in only 68.9% to 95.8% average latency of NVMe/RDMA. However, we also found that target offloading schemes lack flexibility for specific storage device models, leading to abnormal performance degradation. Finally, we provide insight for system architects to select the appropriate scheme based on the workload and storage device features to fully leverage the advantages of target offloading.
- 16 September 2024
- Published by Association for Computing Machinery (ACM)
- p. 14-24
Open AccessPerformance Characterization of SmartNIC NVMe-over-Fabrics Target Offloading
The NVMe-over-Fabrics (NVMe-oF) is gaining popularity in cloud data centers as a remote storage protocol for accessing NVMe storage devices across servers. With the rapid increase in throughput of the NVMe storage devices, the NVMe-oF stack consumes a significant amount of valuable CPU resources. To release these valuable computing power to other tasks, many smartNICs now support NVMe-oF Target offloading. However, this emerging NVMe-oF offloading scheme's performance has not been fully investigated. In this work, we conducted comprehensive evaluations on smartNIC NVMe-oF target offloading and non-offloading NVMe-oF with TCP and RDMA. Our results demonstrate that the target offloading can achieve comparable throughput compared to NVMe/RDMA in most synthetic and real-world application workloads while releasing up to 38.7% CPU resources. Additionally, target offloading benefits from host bypass and peer-to-peer communication between smartNIC and NVMe SSD, resulting in only 68.9% to 95.8% average latency of NVMe/RDMA. However, we also found that target offloading schemes lack flexibility for specific storage device models, leading to abnormal performance degradation. Finally, we provide insight for system architects to select the appropriate scheme based on the workload and storage device features to fully leverage the advantages of target offloading.