Wayne Luk
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
Wayne Luk has a total of 269 co-authors. Here are the top co-authors by number of shared publications:
- Ce Guo
Imperial College London, London, United Kingdom
24 shared publications - Zhiqiang Que
Department of Computing, Imperial College London, London, UK
17 shared publications - Hongxiang Fan
Imperial College London
11 shared publications - Tim Todman
Department of Computing, Imperial College London, London, U.K.
6 shared publications - Xinyu Niu
Corerain Technologies Ltd Shenzhen China
6 shared publications - He Li
Southeast University
5 shared publications
Latest Publications
- Open Access
- 13 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1765-1773
Conformal Trust Control for Learned Sequential Policies under Regime Shift
- 13 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1765-1773
Open AccessConformal Trust Control for Learned Sequential Policies under Regime Shift
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Algorithm and Hardware Co-Design for Efficient Complex-Valued Uncertainty Estimation
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessAlgorithm and Hardware Co-Design for Efficient Complex-Valued Uncertainty Estimation
- 30 June 2026
- Published by Springer Nature
- p. 172-188
Optimizing FPGA-Based Neural Network Ensemble
- 9 June 2026
- Published by Springer Nature
- p. 29-41
VCDF: A Validated Consensus-Driven Framework for Time Series Causal Discovery
- 9 June 2026
- Published by Springer Nature
- p. 29-41
VCDF: A Validated Consensus-Driven Framework for Time Series Causal Discovery
- Top Cited PapersOpen Access
- 27 May 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 19 (2) , 1-35
hls4ml: A Flexible, Open Source Platform for Deep Learning Acceleration on Reconfigurable Hardware
We present hls4ml , a free and open source platform that translates machine learning (ML) models from modern deep learning frameworks into high-level synthesis (HLS) code that can be integrated into full designs for field-programmable gate arrays (FPGAs) or application-specific integrated circuits (ASICs). With its flexible and modular design, hls4ml supports a large number of deep learning frameworks and can target HLS compilers from several vendors, including Vitis HLS, Intel oneAPI and Catapult HLS. Together with a wider eco-system for software-hardware co-design, hls4ml has enabled the acceleration of ML inference in a wide range of commercial and scientific applications where low latency, resource usage, and power consumption are critical. In this article, we describe the structure and functionality of the hls4ml platform. The overarching design considerations for the generated HLS code are discussed, together with selected performance results.
- 27 May 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 19 (2) , 1-35
Top Cited PapersOpen Accesshls4ml: A Flexible, Open Source Platform for Deep Learning Acceleration on Reconfigurable Hardware
We present hls4ml , a free and open source platform that translates machine learning (ML) models from modern deep learning frameworks into high-level synthesis (HLS) code that can be integrated into full designs for field-programmable gate arrays (FPGAs) or application-specific integrated circuits (ASICs). With its flexible and modular design, hls4ml supports a large number of deep learning frameworks and can target HLS compilers from several vendors, including Vitis HLS, Intel oneAPI and Catapult HLS. Together with a wider eco-system for software-hardware co-design, hls4ml has enabled the acceleration of ML inference in a wide range of commercial and scientific applications where low latency, resource usage, and power consumption are critical. In this article, we describe the structure and functionality of the hls4ml platform. The overarching design considerations for the generated HLS code are discussed, together with selected performance results.
- Open Access
- 15 April 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 19 (2) , 1-34
MetaML-Pro: Cross-Stage Design Flow Automation for Efficient Deep Learning Acceleration
This article presents a unified framework for codifying and automating optimization strategies to efficiently deploy deep neural networks (DNNs) on resource-constrained hardware, such as Field-Programmable Gate Arrays (FPGAs), while maintaining high performance, accuracy, and resource efficiency. Deploying DNNs on such platforms involves addressing the significant challenge of balancing performance, resource usage, such as Digital Signal Processing blocks (DSPs) and Look-Up Tables (LUTs), and inference accuracy, which often requires extensive manual effort and domain expertise. Our novel approach addresses two core issues: (i) encoding custom optimization strategies and (ii) enabling cross-stage optimization search. In particular, our proposed framework seamlessly integrates programmatic DNN optimization techniques with high-level synthesis (HLS)-based metaprogramming, leveraging advanced design space exploration (DSE) strategies like Bayesian optimization (BO) to automate both top-down and bottom-up design flows. Hence, we reduce the need for manual intervention and domain expertise. In addition, the framework introduces customizable optimization, transformation, and control blocks to enhance DNN accelerator performance and resource efficiency. We further formalize a cross-stage, constrained BO procedure that couples predicate–action bottom-up feedback (via BRANCH) with FORK/REDUCE order search, enabling automated selection, ordering, and tuning across software and HLS tasks. Experimental results demonstrate up to a 92% DSP and 89% LUT usage reduction for select networks, while preserving accuracy, along with a 15.6-fold reduction in optimization time compared to grid search. These results highlight the potential for automating the generation of resource-efficient DNN accelerator designs with minimal effort, resulting in large resource savings with bounded exploration cost.
- 15 April 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 19 (2) , 1-34
Open AccessMetaML-Pro: Cross-Stage Design Flow Automation for Efficient Deep Learning Acceleration
This article presents a unified framework for codifying and automating optimization strategies to efficiently deploy deep neural networks (DNNs) on resource-constrained hardware, such as Field-Programmable Gate Arrays (FPGAs), while maintaining high performance, accuracy, and resource efficiency. Deploying DNNs on such platforms involves addressing the significant challenge of balancing performance, resource usage, such as Digital Signal Processing blocks (DSPs) and Look-Up Tables (LUTs), and inference accuracy, which often requires extensive manual effort and domain expertise. Our novel approach addresses two core issues: (i) encoding custom optimization strategies and (ii) enabling cross-stage optimization search. In particular, our proposed framework seamlessly integrates programmatic DNN optimization techniques with high-level synthesis (HLS)-based metaprogramming, leveraging advanced design space exploration (DSE) strategies like Bayesian optimization (BO) to automate both top-down and bottom-up design flows. Hence, we reduce the need for manual intervention and domain expertise. In addition, the framework introduces customizable optimization, transformation, and control blocks to enhance DNN accelerator performance and resource efficiency. We further formalize a cross-stage, constrained BO procedure that couples predicate–action bottom-up feedback (via BRANCH) with FORK/REDUCE order search, enabling automated selection, ordering, and tuning across software and HLS tasks. Experimental results demonstrate up to a 92% DSP and 89% LUT usage reduction for select networks, while preserving accuracy, along with a 15.6-fold reduction in optimization time compared to grid search. These results highlight the potential for automating the generation of resource-efficient DNN accelerator designs with minimal effort, resulting in large resource savings with bounded exploration cost.
- Open Access
- 22 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 732-748
FastTTS: Accelerating Test-Time Scaling for Edge LLM Reasoning
Recent advances in reasoning Large Language Models (LLMs) are driving the emergence of agentic AI systems. Edge deployment of LLM agents near end users is increasingly necessary to protect data privacy, enable offline use, and provide responsive interaction with local context. However, strict memory constraints on edge devices limit deployment to smaller LLMs, whose reasoning capabilities are much weaker than those of large cloud models, hindering practical deployment of edge agentic AI. Test-Time Scaling (TTS) offers a promising solution by allocating more compute during inference to enhance the reasoning capability of edge LLMs. However, current TTS methods introduce heavy hardware performance overhead on resource-constrained devices, making them impractical for real applications. To address this challenge, we present FastTTS, a serving system that enables fast and efficient TTS for memory-constrained LLM reasoning. After analyzing common patterns across various TTS methods and identifying their performance bottlenecks, we introduce three novel techniques: i) Speculative Beam Extension, which mitigates system stragglers caused by irregular reasoning paths, ii) Asymmetric Multi-Model Memory Allocation, which dynamically balances memory usage between token generation and reasoning-step verification, and iii) Dynamic Prefix-Aware Scheduling, which optimizes reasoning execution to maximize KV-cache reuse across search paths. FastTTS offers a plug-and-play third-party library on top of vLLM, enabling edge LLMs (łeq 7B) on a single consumer GPU (24 GB) to match cloud-model accuracy and cloud-measured latency. Comprehensive evaluation shows that FastTTS achieves an average 2.2× higher goodput and reduces latency by 38%-68% compared to the vLLM baseline; it pushes the boundaries of low-latency Test-Time Scaling on memory-constrained edge devices and highlights the potential for democratizing agentic AI.
- 22 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 732-748
Open AccessFastTTS: Accelerating Test-Time Scaling for Edge LLM Reasoning
Recent advances in reasoning Large Language Models (LLMs) are driving the emergence of agentic AI systems. Edge deployment of LLM agents near end users is increasingly necessary to protect data privacy, enable offline use, and provide responsive interaction with local context. However, strict memory constraints on edge devices limit deployment to smaller LLMs, whose reasoning capabilities are much weaker than those of large cloud models, hindering practical deployment of edge agentic AI. Test-Time Scaling (TTS) offers a promising solution by allocating more compute during inference to enhance the reasoning capability of edge LLMs. However, current TTS methods introduce heavy hardware performance overhead on resource-constrained devices, making them impractical for real applications. To address this challenge, we present FastTTS, a serving system that enables fast and efficient TTS for memory-constrained LLM reasoning. After analyzing common patterns across various TTS methods and identifying their performance bottlenecks, we introduce three novel techniques: i) Speculative Beam Extension, which mitigates system stragglers caused by irregular reasoning paths, ii) Asymmetric Multi-Model Memory Allocation, which dynamically balances memory usage between token generation and reasoning-step verification, and iii) Dynamic Prefix-Aware Scheduling, which optimizes reasoning execution to maximize KV-cache reuse across search paths. FastTTS offers a plug-and-play third-party library on top of vLLM, enabling edge LLMs (łeq 7B) on a single consumer GPU (24 GB) to match cloud-model accuracy and cloud-measured latency. Comprehensive evaluation shows that FastTTS achieves an average 2.2× higher goodput and reduces latency by 38%-68% compared to the vLLM baseline; it pushes the boundaries of low-latency Test-Time Scaling on memory-constrained edge devices and highlights the potential for democratizing agentic AI.
- Top Cited PapersOpen Access
- 6 March 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 19 (1) , 1-27
da4ml: Distributed Arithmetic for Real-time Neural Networks on FPGAs
Neural networks with a latency requirement on the order of microseconds, like the ones used at the CERN Large Hadron Collider, are typically deployed on FPGAs fully unrolled and pipelined. A bottleneck for the deployment of such neural networks is area utilization, which is directly related to the required constant matrix-vector multiplication (CMVM) operations. In this work, we propose an efficient algorithm for implementing CMVM operations with distributed arithmetic on FPGAs that simultaneously optimizes for area consumption and latency. The algorithm achieves resource reduction similar to state-of-the-art algorithms while being significantly faster to compute. The proposed algorithm is open sourced and integrated into the hls4ml library, a free and open source library for running real-time neural network inference on FPGAs. We show that the proposed algorithm can reduce on-chip resources by up to a third for realistic, highly quantized neural networks while simultaneously reducing latency, enabling the implementation of previously infeasible networks.
- 6 March 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 19 (1) , 1-27
Top Cited PapersOpen Accessda4ml: Distributed Arithmetic for Real-time Neural Networks on FPGAs
Neural networks with a latency requirement on the order of microseconds, like the ones used at the CERN Large Hadron Collider, are typically deployed on FPGAs fully unrolled and pipelined. A bottleneck for the deployment of such neural networks is area utilization, which is directly related to the required constant matrix-vector multiplication (CMVM) operations. In this work, we propose an efficient algorithm for implementing CMVM operations with distributed arithmetic on FPGAs that simultaneously optimizes for area consumption and latency. The algorithm achieves resource reduction similar to state-of-the-art algorithms while being significantly faster to compute. The proposed algorithm is open sourced and integrated into the hls4ml library, a free and open source library for running real-time neural network inference on FPGAs. We show that the proposed algorithm can reduce on-chip resources by up to a third for realistic, highly quantized neural networks while simultaneously reducing latency, enabling the implementation of previously infeasible networks.
- Open Access
- 21 February 2026
- Published by Association for Computing Machinery (ACM)
- p. 186
CODESCA: Co-Design for Spectral Clustering Acceleration
- 21 February 2026
- Published by Association for Computing Machinery (ACM)
- p. 186
Open AccessCODESCA: Co-Design for Spectral Clustering Acceleration
- Open Access
- 21 February 2026
- Published by Association for Computing Machinery (ACM)
- p. 79-91
HGQ: High Granularity Quantization for Real-time Neural Networks on FPGAs
- 21 February 2026
- Published by Association for Computing Machinery (ACM)
- p. 79-91
Open AccessHGQ: High Granularity Quantization for Real-time Neural Networks on FPGAs