PhD Candidate, HKUST ECE / ACCESS
Kunming SHAO 邵堃明
I am a PhD candidate at The Hong Kong University of Science and Technology (HKUST) and the AI Chip Center for Emerging Smart Systems (ACCESS), supervised by Prof. Chi-Ying TSUI and Prof. Tim Kwang-Ting CHENG. My research connects digital compute-in-memory circuits, AI accelerator architecture, compiler automation, efficient ML systems, edge LLM agents, and quantization.
I expect to graduate in Summer 2027 and am actively exploring postdoctoral opportunities in AI hardware, computing-in-memory, and efficient intelligence systems.
HKUST ECE / ACCESS
Excellent graduation thesis
Fellowship and academic awards
Research Focus
Hardware-efficient intelligence from circuit to system
Digital CIM and ReRAM
Macro architecture, robust in-memory computation, hybrid SRAM/ReRAM cells, and design automation for digital compute-in-memory.
AI accelerators
Energy-efficient accelerators for DNNs, SNNs, Transformers, FlashAttention, stochastic/approximate computing, and FP8 computation.
ML systems and agents
Edge RAG, wearable medical LLM agents, retrieval acceleration, quantization, and hardware-aware deployment of efficient models.
Selected Threads
Recent publication directions
ReRAM-coupled digital CIM accelerator for FlashAttention dataflow.
DATE'26 DS-CIMDigital stochastic CIM with accurate OR-accumulation for edge AI models.
ISLPED'25 DIRC-RAGHigh-density digital In-ReRAM computation for edge RAG retrieval.
DATE'25 SynDCIMPerformance-aware DCIM compiler with multi-spec subcircuit synthesis.
News
Latest updates
RedBird Award: I received the HKUST RedBird Academic Excellence Award.
ESSERC'26: My first-authored paper SwiftCIM: a 55nm 23.2μJ/Token L-0.5 ReRAM Coupled Digital CIM Accelerator with Fully-Fused Multi-Head Attention Dataflow for FlashAttention was accepted.
TVLSI: The paper I led and co-first-authored, Balancing FP8 Computation Accuracy and Efficiency on Digital CIM via Shift-Aware On-the-fly Aligned-Mantissa Bitwidth Prediction, was accepted.
US Patent: Our co-authored patent on a hybrid computing-in-memory device and multi-level sensing method was approved and published.
Chinese Patent: Our co-authored patent on a hybrid CIM device and multi-level data-bit sensing method was approved and published.
DATE'26: My first-authored paper DS-CIM: Digital Stochastic Computing-In-Memory Featuring Accurate OR-Accumulation via Sample Region Remapping for Edge AI Models was accepted.
A-SSCC'25: Our co-authored paper Lemem: A 179.8TFLOPS/W, 24.21TFLOPS Learning-In-Memory Processor with Layer-Fused Forward/Backward Pipeline for Edge DNN/SNN Training/Inference was accepted.
BioCAS'25: The paper I led and co-first-authored, A Memory-Efficient Retrieval Architecture for RAG-Enabled Wearable Medical LLMs-Agents, was accepted.
TCAD: Our co-authored paper Configurable Dataflow and Adaptive Mapping Optimization for Hybrid ReRAM and SRAM Compute-in-Memory Accelerator was accepted.
CASS Travel Grant: I received the IEEE CASS Student Travel Grant.
ISLPED'25: My first-authored paper DIRC-RAG: Accelerating Edge RAG with Robust High-Density and High-Loading-Bandwidth Digital In-ReRAM Computation was accepted.
DAC'25 WIP: My work-in-progress poster on AI accelerators based on approximate computing was accepted.
CICC'25: Our co-first-authored paper E-NPU: A 34~126nJ/Class Event-Driven Adaptive Neural SoC with Signal-Dynamics-Aware Feature Clustering and Multi-Model In-Memory Inference/Training for Personalized Medical Wearables was accepted.
ISCAS'25: Our co-first-authored paper A Flexible Precision Scaling Deep Neural Network Accelerator with Efficient Weight Combination was accepted.
PQE: I passed the PhD Qualification Exam and continued as a PhD candidate.
DATE'25: My first-authored paper SynDCIM: A Performance-Aware Digital Computing-in-Memory Compiler with Multi-Spec-Oriented Subcircuit Synthesis was accepted.
ICCAD'24: Our co-authored paper ReSCIM: Variation-Resilient High Weight-Loading Bandwidth In-Memory Computation Based on Fine-Grained Hybrid Integration of Multi-Level ReRAM and SRAM Cells was accepted.
Thesis: My BEng thesis, Digital Compute-In-Memory Automatic Design Methodology, was selected as an excellent graduation project at SCUT.
HKPFS & RedBird: I received the Hong Kong PhD Fellowship and HKUST RedBird Award.
DAC'23: Our co-authored paper AutoDCIM: An Automated Digital CIM Compiler was accepted.
Collaborations
Research group and joint projects
I coordinate a multi-institution collaboration across HKUST, SCUT, Westlake University, SYSU, WHU, and other partners on in-memory computation, approximate computing, efficient algorithms, and emerging non-volatile memories. The projects below highlight first/co-first author and corresponding-author roles.
Kunming Shao and Fengshi Tian, HKUST.
Liang Zhao, SCUT, and Kunming Shao, HKUST.
Fengshi Tian, HKUST; Jinbo Chen, Westlake; Kunming Shao, HKUST.
Kunming Shao, HKUST; Zhipeng Liao, Westlake; Jiangnan Yu and Xiaomeng Wang, HKUST.
Zhipeng Liao, Westlake, and Kunming Shao, HKUST.
Kunming Shao, Liang Zhao, Xiaomeng Wang, and collaborators.