Huaxia Xia
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Huaxia Xia has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Chen Zuoning
Chinese Academy of Engineering, Wuxi, China
1 shared publication - Feng Ren
Computer Science and Technology, Tsinghua University
1 shared publication - Kang Chen
Tsinghua University, Beijing, China
1 shared publication - Mingxing Zhang
Tsinghua University, Beijing, China
1 shared publication - Qing-Shan Jia
Center for Intelligent and Networked Systems (CFINS), Department of Automation, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua University, Beijing, China
1 shared publication - Tao Zhang
Tsinghua University
1 shared publication
Latest Publications
- Open Access
- 17 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 351-367
Scaling Up Memory Disaggregated Applications with SMART
Recent developments in RDMA networks are leading to the trend of memory disaggregation. However, the performance of each compute node is still limited by the network, especially when it needs to perform a large number of concurrent fine-grained remote accesses. According to our evaluations, existing IOPS-bound disaggregated applications do not scale well beyond 32 cores, and therefore do not take full advantage of today's many-core machines. After an in-depth analysis of the internal architecture of RNIC, we found three major scale-up bottlenecks that limit the throughput of today's disaggregated applications: (1) implicit contention of doorbell registers, (2) cache trashing caused by excessive outstanding work requests, and (3) wasted IOPS from unsuccessful CAS retries. However, the solutions to these problems involve many low-level details that are not familiar to application developers. To ease the burden on developers, we propose Smart, an RDMA programming framework that hides the above details by providing an interface similar to one-sided RDMA verbs. We take 44 and 16 lines of code to refactor the state-of-the-art disaggregated hash table (RACE) and persistent transaction processing system (FORD) with Smart, improving their throughput by up to 132.4× and 5.2×, respectively. We have also refactored Sherman (a recent disaggregated B+Tree) with Smart and an additional speculative lookup optimization (48 lines of code changed), which changes its memory access pattern from bandwidth-bound to IOPS-bound and leads to a speedup of 2.0×. Smart is publicly available at https://github.com/madsys-dev/smart.
- 17 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 351-367
Open AccessScaling Up Memory Disaggregated Applications with SMART
Recent developments in RDMA networks are leading to the trend of memory disaggregation. However, the performance of each compute node is still limited by the network, especially when it needs to perform a large number of concurrent fine-grained remote accesses. According to our evaluations, existing IOPS-bound disaggregated applications do not scale well beyond 32 cores, and therefore do not take full advantage of today's many-core machines. After an in-depth analysis of the internal architecture of RNIC, we found three major scale-up bottlenecks that limit the throughput of today's disaggregated applications: (1) implicit contention of doorbell registers, (2) cache trashing caused by excessive outstanding work requests, and (3) wasted IOPS from unsuccessful CAS retries. However, the solutions to these problems involve many low-level details that are not familiar to application developers. To ease the burden on developers, we propose Smart, an RDMA programming framework that hides the above details by providing an interface similar to one-sided RDMA verbs. We take 44 and 16 lines of code to refactor the state-of-the-art disaggregated hash table (RACE) and persistent transaction processing system (FORD) with Smart, improving their throughput by up to 132.4× and 5.2×, respectively. We have also refactored Sherman (a recent disaggregated B+Tree) with Smart and an additional speculative lookup optimization (48 lines of code changed), which changes its memory access pattern from bandwidth-bound to IOPS-bound and leads to a speedup of 2.0×. Smart is publicly available at https://github.com/madsys-dev/smart.
- 23 August 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1967-1972
Learning-Based Safety-Critical Motion Planning with Input-to-State Barrier Certificate
Motion planning in an effective and safe manner is a critical yet challenging task for autonomous driving. Learning-based framework as a new fashion in simulation and optimization has great potentials to develop a time-efficient navigation policy. To design a controller that addresses safety with explicit formulation, we incorporate control barrier function approach to generate a constrained optimization problem. The proposed method with uncertainty analysis helps to deal with disturbance towards motion planning task. Simulation results show that the algorithm produces improvement in the learning process and the adaptation to safety and performance.
- 23 August 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1967-1972
Learning-Based Safety-Critical Motion Planning with Input-to-State Barrier Certificate
Motion planning in an effective and safe manner is a critical yet challenging task for autonomous driving. Learning-based framework as a new fashion in simulation and optimization has great potentials to develop a time-efficient navigation policy. To design a controller that addresses safety with explicit formulation, we incorporate control barrier function approach to generate a constrained optimization problem. The proposed method with uncertainty analysis helps to deal with disturbance towards motion planning task. Simulation results show that the algorithm produces improvement in the learning process and the adaptation to safety and performance.