Mingyu Wu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Mingyu Wu has a total of 56 co-authors. Here are the top co-authors by number of shared publications:
- Haibo Chen
Shanghai Jiao Tong University, Shanghai, China
12 shared publications - Binyu Zang
Shanghai Jiao Tong University, Shanghai, China
7 shared publications - Ziming Zhao
Institute of Parallel and Distributed Systems, SEIEE, Shanghai Jiao Tong University, Shanghai, China
3 shared publications - Fangming Lu
Institute of Parallel and Distributed Systems, Shanghai Jiao Tong University
2 shared publications - Jinze Si
Shanghai Jiao Tong University, Shanghai, China
2 shared publications - Rong Chen
Shanghai Jiao Tong University, Shanghai, China
2 shared publications
Latest Publications
- Open Access
- 14 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1290-1304
EvoFuzz: IR-Aware Fuzzing for JIT Compilers through the Lens of Node Evolution
- 14 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1290-1304
Open AccessEvoFuzz: IR-Aware Fuzzing for JIT Compilers through the Lens of Node Evolution
- Open Access
- 14 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1862-1877
Understanding and Detecting Performance Regressions in Modern Open-Source Server Applications
- 14 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1862-1877
Open AccessUnderstanding and Detecting Performance Regressions in Modern Open-Source Server Applications
- Open Access
- 14 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 334-351
Chameleon: Adaptive Memory Overcommitment for Disaggregated Virtual Machines
- 14 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 334-351
Open AccessChameleon: Adaptive Memory Overcommitment for Disaggregated Virtual Machines
- 4 November 2026
- Published by Springer Nature in Nature
An AI-native experimental laboratory for autonomous biomolecular operations
- 4 November 2026
- Published by Springer Nature in Nature
An AI-native experimental laboratory for autonomous biomolecular operations
- Open Access
- 17 February 2026
- Published by Association for Computing Machinery (ACM)
- p. 295-301
miniK8s: A Pedagogical Cloud-Native System
- 17 February 2026
- Published by Association for Computing Machinery (ACM)
- p. 295-301
Open AccessminiK8s: A Pedagogical Cloud-Native System
- Open Access
- 16 December 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 22 (4) , 1-26
Object-Aware Memory Compression for Smartphones
Memory compression is vitally important in controlling the memory consumption of smartphones. Unfortunately, mainstream compression schemes like zRAM do not work well with garbage collection (GC) in managed language runtimes. When GC is triggered, it traces objects in the heap and thus causes frequent swapping and repetitive memory compression/decompression, which introduces significant performance overhead and user experience degradation. To this end, we propose an object-aware compressed memory scheme named oaRAM , which allows GC algorithms to directly interpret compressed memory and thus eliminates the need for decompression during GC. Furthermore, we build swap-less memory reclamation by further bridging the semantic gap between the memory compression module in OS and the garbage collector in the language runtime. On the OS side, we provide an object-aware memory compression module, which leverages object semantics from language runtimes to compress and manage data in the kernel. On the language runtime side, we propose shadow heap, which allows GC threads to directly access compressed contents in the kernel. We have implemented oaRAM on Android, and the evaluation results show that it significantly reduces the number of swapping operations under memory-hungry scenarios, which results in improvements in application performance, the system’s application caching capability and device lifespan.
- 16 December 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 22 (4) , 1-26
Open AccessObject-Aware Memory Compression for Smartphones
Memory compression is vitally important in controlling the memory consumption of smartphones. Unfortunately, mainstream compression schemes like zRAM do not work well with garbage collection (GC) in managed language runtimes. When GC is triggered, it traces objects in the heap and thus causes frequent swapping and repetitive memory compression/decompression, which introduces significant performance overhead and user experience degradation. To this end, we propose an object-aware compressed memory scheme named oaRAM , which allows GC algorithms to directly interpret compressed memory and thus eliminates the need for decompression during GC. Furthermore, we build swap-less memory reclamation by further bridging the semantic gap between the memory compression module in OS and the garbage collector in the language runtime. On the OS side, we provide an object-aware memory compression module, which leverages object semantics from language runtimes to compress and manage data in the kernel. On the language runtime side, we propose shadow heap, which allows GC threads to directly access compressed contents in the kernel. We have implemented oaRAM on Android, and the evaluation results show that it significantly reduces the number of swapping operations under memory-hungry scenarios, which results in improvements in application performance, the system’s application caching capability and device lifespan.
- Open Access
- 6 November 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Computer Systems
- Vol. 43 (4) , 1-32
Towards Serialization/Deserialization-free State Transfer in Serverless Workflows
Serialization and deserialization dominate the state transfer time of serverless workflows, leading to substantial performance penalties when executing various serverless workflow applications. We identify the key reason for serialization and deserialization as a lack of ability to efficiently access the (remote) memory of another function. To this end, we propose RMMap , an OS primitive for remote memory map, which allows a serverless function to directly access the memory of another function, even if it is located remotely. RMMap is the first to completely eliminate serialization and deserialization overhead when transferring states between any pairs of functions in (unmodified) serverless workflows. To make remote memory map efficient and feasible, we co-design it with modern networking (RDMA), OS, language runtime, and serverless platform. Evaluations using real-world serverless workloads show that integrating RMMap with Knative reduces the serverless workflow execution time on Knative by up to 2.6× and improves resource utilizations by 86.3%.
- 6 November 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Computer Systems
- Vol. 43 (4) , 1-32
Open AccessTowards Serialization/Deserialization-free State Transfer in Serverless Workflows
Serialization and deserialization dominate the state transfer time of serverless workflows, leading to substantial performance penalties when executing various serverless workflow applications. We identify the key reason for serialization and deserialization as a lack of ability to efficiently access the (remote) memory of another function. To this end, we propose RMMap , an OS primitive for remote memory map, which allows a serverless function to directly access the memory of another function, even if it is located remotely. RMMap is the first to completely eliminate serialization and deserialization overhead when transferring states between any pairs of functions in (unmodified) serverless workflows. To make remote memory map efficient and feasible, we co-design it with modern networking (RDMA), OS, language runtime, and serverless platform. Evaluations using real-world serverless workloads show that integrating RMMap with Knative reduces the serverless workflow execution time on Knative by up to 2.6× and improves resource utilizations by 86.3%.
- Open Access
- 22 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 281-297
Characterization and Reclamation of Frozen Garbage in Managed FaaS Workloads
- 22 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 281-297
Open AccessCharacterization and Reclamation of Frozen Garbage in Managed FaaS Workloads
- Open Access
- 22 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 132-147
Serialization/Deserialization-free State Transfer in Serverless Workflows
- 22 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 132-147
Open AccessSerialization/Deserialization-free State Transfer in Serverless Workflows
- Open Access
- 22 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 1160-1174
Jade: A High-throughput Concurrent Copying Garbage Collector
Garbage collection (GC) pauses are a notorious issue threatening the latency of applications. To mitigate this problem, state-of-the-art concurrent copying collectors allow GC threads to run simultaneously with application threads (mutators) in nearly all GC phases. However, the design of concurrent copying collectors does not always lead to low application latency. To this end, this work studies the behaviors of mainstream concurrent copying collectors in OpenJDK and mainly focuses on long application pauses under heavy workloads. By analyzing the design of those collectors, this work uncovers that lengthy pre-reclamation cycles (including GC phases before actual memory release), high GC frequency, and large metadata maintenance overhead are major factors for long pauses. Therefore, this work proposes Jade, a concurrent copying collector aiming to achieve both short pauses and high GC efficiency. Compared with existing collectors, Jade provides a group-wise collection mechanism to shorten pre-reclamation cycles while controlling GC frequency. It also embraces a generational heap layout and a single-phase algorithm to maximize young GC's throughput. The evaluation results on representative latency-critical applications show that Jade can reach sub-millisecond-level pauses even under heavy workloads and significantly improve applications' peak throughput compared with state-of-the-art concurrent collectors.
- 22 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 1160-1174
Open AccessJade: A High-throughput Concurrent Copying Garbage Collector
Garbage collection (GC) pauses are a notorious issue threatening the latency of applications. To mitigate this problem, state-of-the-art concurrent copying collectors allow GC threads to run simultaneously with application threads (mutators) in nearly all GC phases. However, the design of concurrent copying collectors does not always lead to low application latency. To this end, this work studies the behaviors of mainstream concurrent copying collectors in OpenJDK and mainly focuses on long application pauses under heavy workloads. By analyzing the design of those collectors, this work uncovers that lengthy pre-reclamation cycles (including GC phases before actual memory release), high GC frequency, and large metadata maintenance overhead are major factors for long pauses. Therefore, this work proposes Jade, a concurrent copying collector aiming to achieve both short pauses and high GC efficiency. Compared with existing collectors, Jade provides a group-wise collection mechanism to shorten pre-reclamation cycles while controlling GC frequency. It also embraces a generational heap layout and a single-phase algorithm to maximize young GC's throughput. The evaluation results on representative latency-critical applications show that Jade can reach sub-millisecond-level pauses even under heavy workloads and significantly improve applications' peak throughput compared with state-of-the-art concurrent collectors.