Xuechen Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Xuechen Zhang has a total of 80 co-authors. Here are the top co-authors by number of shared publications:
- Shuibing He
Zhejiang University
14 shared publications - Gang Chen
Zhejiang University
7 shared publications - Xian-He Sun
Illinois Institute of Technology
7 shared publications - Ping Chen
Zhejiang University
5 shared publications - Siling Yang
Zhejiang University
5 shared publications - Weijian Chen
Zhejiang University
5 shared publications
Latest Publications
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 75 (12) , 4091-4105
HyperInfer: Accelerating Large Language Model Inference via Importance-Informed Prefix-KV Caching and Prefetching
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 75 (12) , 4091-4105
HyperInfer: Accelerating Large Language Model Inference via Importance-Informed Prefix-KV Caching and Prefetching
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 75 (11) , 4010-4024
SIFMap: Parallel CNN Inference Using Sub-Input-Feature Maps in PIM
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 75 (11) , 4010-4024
SIFMap: Parallel CNN Inference Using Sub-Input-Feature Maps in PIM
- Open Access
- 27 September 2026
- Published by Association for Computing Machinery (ACM)
- p. 523-532
NeuroPrefetcher: Storage-Aware Sparse LLM Inference via Delta Prefetching
- 27 September 2026
- Published by Association for Computing Machinery (ACM)
- p. 523-532
Open AccessNeuroPrefetcher: Storage-Aware Sparse LLM Inference via Delta Prefetching
- Open Access
- 13 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 642-650
Scalable Quantum Circuit Simulation via Circuit Cutting and FPGA Acceleration
This paper presents a scalable quantum circuit simulation framework that integrates circuit cutting with specialized hardware acceleration. While circuit cutting mitigates the exponential memory growth of classical simulation, it introduces substantial reconstruction overhead that becomes a key performance bottleneck. To address this, we propose a specialized hardware architecture optimized for massively parallel computations. Through algorithm–hardware co-design, the proposed platform significantly accelerates reconstruction while maintaining high numerical accuracy (error bound: 3.05 × 10− 5) and efficient resource utilization. Case studies demonstrate nanosecond-scale performance enabled by spatial parallelism, deep pipelining, and improved memory locality. Although validated on moderate-scale circuits, the framework establishes a scalable foundation for high-performance quantum circuit simulation through tight integration of circuit cutting and architecture-aware design.
- 13 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 642-650
Open AccessScalable Quantum Circuit Simulation via Circuit Cutting and FPGA Acceleration
This paper presents a scalable quantum circuit simulation framework that integrates circuit cutting with specialized hardware acceleration. While circuit cutting mitigates the exponential memory growth of classical simulation, it introduces substantial reconstruction overhead that becomes a key performance bottleneck. To address this, we propose a specialized hardware architecture optimized for massively parallel computations. Through algorithm–hardware co-design, the proposed platform significantly accelerates reconstruction while maintaining high numerical accuracy (error bound: 3.05 × 10− 5) and efficient resource utilization. Case studies demonstrate nanosecond-scale performance enabled by spatial parallelism, deep pipelining, and improved memory locality. Although validated on moderate-scale circuits, the framework establishes a scalable foundation for high-performance quantum circuit simulation through tight integration of circuit cutting and architecture-aware design.
- Open Access
- 8 July 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
NVREC: A High-Performance Recoverable Recommendation System for Non-Volatile Memory
Current recommendation systems are designed with the assumption that computer memory is volatile. Thus, they periodically execute checkpointing operations to save model weights and embeddings to slower solid-state drives (SSDs) for failure recovery. Such designs cannot leverage the unique characteristics of non-volatile memory (NVM). This paper proposes a new recommendation model training framework called NVrec , which uses a hybrid memory system consisting of DRAM and NVM to support computing and in-memory checkpointing. NVrec proposes three recommendation-oriented techniques. (1) It partitions embeddings so that only popular vectors are placed in DRAM to hide NVM-induced latency. (2) It maintains a dual-version scheme for in-memory checkpointing: each vector in NVM alternates between a persistent version for crash recovery and a runtime version for training updates. (3) It employs fine-grained parallelism to hide the checkpointing overhead of model parameters in DRAM and GPU. Our experiments demonstrate that NVrec improves recommendation-model training throughput by 1.05 × -6.85 × and the checkpointing process by more than 5 × compared to existing models when evaluated on the synthetic and real-world datasets.
- 8 July 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
Open AccessNVREC: A High-Performance Recoverable Recommendation System for Non-Volatile Memory
Current recommendation systems are designed with the assumption that computer memory is volatile. Thus, they periodically execute checkpointing operations to save model weights and embeddings to slower solid-state drives (SSDs) for failure recovery. Such designs cannot leverage the unique characteristics of non-volatile memory (NVM). This paper proposes a new recommendation model training framework called NVrec , which uses a hybrid memory system consisting of DRAM and NVM to support computing and in-memory checkpointing. NVrec proposes three recommendation-oriented techniques. (1) It partitions embeddings so that only popular vectors are placed in DRAM to hide NVM-induced latency. (2) It maintains a dual-version scheme for in-memory checkpointing: each vector in NVM alternates between a persistent version for crash recovery and a runtime version for training updates. (3) It employs fine-grained parallelism to hide the checkpointing overhead of model parameters in DRAM and GPU. Our experiments demonstrate that NVrec improves recommendation-model training throughput by 1.05 × -6.85 × and the checkpointing process by more than 5 × compared to existing models when evaluated on the synthetic and real-world datasets.
- 15 April 2026
- Published by Elsevier in Journal of Parallel and Distributed Computing
- Vol. 213
PMIO: Optimizing parallel collective I/O accesses through node-local persistent memory
- 15 April 2026
- Published by Elsevier in Journal of Parallel and Distributed Computing
- Vol. 213
PMIO: Optimizing parallel collective I/O accesses through node-local persistent memory
- Open Access
CRouting: Reducing Expensive Distance Calls in Graph-Based Approximate Nearest Neighbor Search
Approximate nearest neighbor search (ANNS) is a crucial problem in information retrieval and AI applications. Recently, there has been a surge of interest in graph-based ANNS algorithms due to their superior efficiency and accuracy. However, the repeated computation of distances in high-dimensional spaces constitutes the primary time cost of graph-based methods. To accelerate the search, we propose a novel routing strategy named CRouting, which bypasses unnecessary distance computations by exploiting the angle distributions of high-dimensional vectors. CRouting is designed as a plugin to optimize existing graph-based search with minimal code modifications. Our experiments show that CRouting reduces the number of distance computations by up to 41.5% and boosts queries per second by up to 1.48$\times$ on two predominant graph indexes, HNSW and NSG. Code is publicly available at https://github.com/ISCS-ZJU/CRouting.
Open AccessCRouting: Reducing Expensive Distance Calls in Graph-Based Approximate Nearest Neighbor Search
Approximate nearest neighbor search (ANNS) is a crucial problem in information retrieval and AI applications. Recently, there has been a surge of interest in graph-based ANNS algorithms due to their superior efficiency and accuracy. However, the repeated computation of distances in high-dimensional spaces constitutes the primary time cost of graph-based methods. To accelerate the search, we propose a novel routing strategy named CRouting, which bypasses unnecessary distance computations by exploiting the angle distributions of high-dimensional vectors. CRouting is designed as a plugin to optimize existing graph-based search with minimal code modifications. Our experiments show that CRouting reduces the number of distance computations by up to 41.5% and boosts queries per second by up to 1.48$\times$ on two predominant graph indexes, HNSW and NSG. Code is publicly available at https://github.com/ISCS-ZJU/CRouting.
- 23 August 2025
- Published by Springer Nature
- p. 264-277
iAug: Accelerating Augmentation with Importance Sampling in Deep Neural Network Training
- 23 August 2025
- Published by Springer Nature
- p. 264-277
iAug: Accelerating Augmentation with Importance Sampling in Deep Neural Network Training
- 9 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 74 (8) , 2649-2662
ImPACT: Importance-Informed Prefetching and Caching for I/O-Bound DNN Training
Fetching large amounts of DNN training data from storage systems causes high I/O latency and GPU stalls. Importance sampling can reduce data processing on GPUs while maintaining model accuracy, but current frameworks lack a prefetching and caching layer to optimize data fetches and cache management based on sample importance. This leads to unnecessary fetches, poor cache hit ratios, and random I/Os. We present ImPACT, an importance-informed prefetching and caching system, to accelerate I/O-bound DNN training. First, we propose an importance-informed prefetching technique to reduce the prefetching of unimportant data. Then, we introduce an importance-aware caching layer, partitioned into two regions: H-cache and L-cache, which store samples of high importance and low importance respectively. Rather than using recency or frequency, we manage data items in H-cache according to their corresponding sample importance. When there is a cache miss in L-cache, we use sample substitutability and dynamic packaging to improve the cache hit ratio and reduce the number of random I/Os. Our experimental results show that ImPACT has a negligible impact on training accuracy while speeding up DNN training by up to 3.5$ \times $ compared to state-of-the-art prefetching and caching systems.- 9 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 74 (8) , 2649-2662
ImPACT: Importance-Informed Prefetching and Caching for I/O-Bound DNN Training
Fetching large amounts of DNN training data from storage systems causes high I/O latency and GPU stalls. Importance sampling can reduce data processing on GPUs while maintaining model accuracy, but current frameworks lack a prefetching and caching layer to optimize data fetches and cache management based on sample importance. This leads to unnecessary fetches, poor cache hit ratios, and random I/Os. We present ImPACT, an importance-informed prefetching and caching system, to accelerate I/O-bound DNN training. First, we propose an importance-informed prefetching technique to reduce the prefetching of unimportant data. Then, we introduce an importance-aware caching layer, partitioned into two regions: H-cache and L-cache, which store samples of high importance and low importance respectively. Rather than using recency or frequency, we manage data items in H-cache according to their corresponding sample importance. When there is a cache miss in L-cache, we use sample substitutability and dynamic packaging to improve the cache hit ratio and reduce the number of random I/Os. Our experimental results show that ImPACT has a negligible impact on training accuracy while speeding up DNN training by up to 3.5$ \times $ compared to state-of-the-art prefetching and caching systems.- 5 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 74 (8) , 2552-2566
Advanced Maximal Biclique Enumeration on GPUs Using Bitmaps
Maximal biclique enumeration (MBE) in bipartite graphs is an important problem in data mining with many real-world applications. Parallel MBE algorithms for GPUs are needed for MBE acceleration leveraging its many computing cores. However, enumerating maximal bicliques using GPUs has three main challenges including large memory requirement, thread divergence, and load imbalance. In this paper, we propose GMBE+, an advanced GPU solution for the MBE problem. To overcome the challenges, we design (1) a node-reuse approach to reduce GPU memory usage with advanced node pruning, (2) a bitmap-based set intersection approach to minimize thread divergence, and (3) a load-aware task scheduling framework to achieve load balance among threads within GPU warps, facilitated by a novel set union approach. Our experiments reveal that GMBE+ is 1.2× faster than the latest GPU-based MBE algorithm GMBE on average when running on the same NVIDIA A100 GPU.
- 5 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 74 (8) , 2552-2566
Advanced Maximal Biclique Enumeration on GPUs Using Bitmaps
Maximal biclique enumeration (MBE) in bipartite graphs is an important problem in data mining with many real-world applications. Parallel MBE algorithms for GPUs are needed for MBE acceleration leveraging its many computing cores. However, enumerating maximal bicliques using GPUs has three main challenges including large memory requirement, thread divergence, and load imbalance. In this paper, we propose GMBE+, an advanced GPU solution for the MBE problem. To overcome the challenges, we design (1) a node-reuse approach to reduce GPU memory usage with advanced node pruning, (2) a bitmap-based set intersection approach to minimize thread divergence, and (3) a load-aware task scheduling framework to achieve load balance among threads within GPU warps, facilitated by a novel set union approach. Our experiments reveal that GMBE+ is 1.2× faster than the latest GPU-based MBE algorithm GMBE on average when running on the same NVIDIA A100 GPU.