Shengchao Liu | Computer Science | Research Excellence Award

Published on by Invention Awards

Dr. Shengchao Liu | Computer Science | Research Excellence Award

The Chinese University | Hong Kong

Shengchao Liu is a tenure-track Assistant Professor in the Department of Computer Science and Engineering at The Chinese University of Hong Kong, whose research lies at the intersection of machine learning, geometry, and scientific discovery. His work focuses on developing foundation models and physics-inspired learning frameworks for molecules, proteins, and materials, with the long-term goal of accelerating discovery in chemistry, biology, and materials science. By integrating multi-modal data, symmetry principles, and domain knowledge, his research bridges theoretical advances in AI with real-world experimental impact. A central theme of Dr. Liu’s research is geometric and symmetry-informed representation learning. He has pioneered group-equivariant and manifold-constrained generative models that respect the underlying physical laws of molecular and material systems. His contributions include SE(3)-invariant pretraining methods, group-symmetric stochastic differential equation models, and rigid flow matching techniques, which have significantly improved the fidelity and interpretability of molecular generation and dynamics modeling. These methods form a unifying framework for learning across molecules, proteins, and crystalline materials, as demonstrated in his influential works at ICLR, ICML, NeurIPS, and AISTATS. Dr. Liu’s work is deeply collaborative and interdisciplinary. He has worked closely with leading researchers across academia and industry, including Mila, UC Berkeley, NVIDIA Research, and national laboratories. As a Principal Investigator, he has led NERSC-supported projects on foundation models for material discovery, leveraging large-scale GPU resources to push the frontier of generative AI for science. His research has also contributed widely used open-source resources, including geometric graph learning benchmarks and toolkits adopted by the broader AI-for-science community.

Citation Metrics (Google Scholar)

4000
3000
2000
1000
  500
  100
      0

Citations
3510

Documents
40

h-index
22

Citations

Documents

h-index

View Google Scholar Profile

Featured Publications


Pre-training Molecular Graph Representation with 3D Geometry

– International Conference on Learning Representations , 2021 | Cited by 574


N-Gram Graph: Simple Unsupervised Representation for Graphs, with Applications to Molecules

– Advances in Neural Information Processing Systems, 2019 | Cited by 295


A text-guided protein design framework

– Nature Machine Intelligence, 2025 | Cited by 225

 

Mr. Jiandong Ma | Computer Science | Best Researcher Award

Published on by Invention Awards

Mr. Jiandong Ma | Computer Science | Best Researcher Award

Chinese Academy of Sciences, China.

Jiandong Ma is a talented engineer specializing in signal and information processing, with expertise in FPGA development and network engineering. He is currently working on cutting-edge projects in the field of RDMA NIC and multipath SD-WAN, leading teams to achieve groundbreaking advancements in data transmission and network performance. With several patents to his name, Jiandong has demonstrated a strong commitment to innovation in network technologies. His contributions have positioned him as a promising figure in the area of network engineering.

Profile

Orcid

Education 🎓

Jiandong Ma completed his Ph.D. in Signal and Information Processing at the University of Chinese Academy of Sciences (UCAS), where he specialized in electronic, electrical, and communication engineering. His doctoral research was conducted under the National Network New Media Engineering Research Center at the Institute of Acoustics, CAS. His academic journey is characterized by his pursuit of innovative solutions in the field of network engineering, focusing on cutting-edge technologies such as RDMA NICs and network performance optimization.

Prior to his Ph.D., Jiandong earned his Bachelor's degree in Network Engineering from the University of Electronic Science and Technology of China (UESTC). As part of the Yingcai Honors College of UESTC, he demonstrated exceptional academic performance, ranking highly in his class. His undergraduate studies laid the foundation for his deep interest in signal processing and network systems, which would later drive his successful research career.

Experience 🛠

FPGA - Out-of-Order (OOO) RDMA NIC (Present)
Team Leader
Jiandong led a team to develop a high-speed RDMA NIC supporting out-of-order packets via Xilinx ERNIC IP, achieving improved WQE transmission performance under multipath scenarios. This innovation has been applied for patents.

FPGA - Packet Reordering and Deduplication System (Completed)
Team Leader
Developed a packet reordering and deduplication system for SD-WAN, reducing space usage by 20% and achieving a throughput of 100 Gbps. The project is patented.

FPGA - Deep Flow Table (Completed)
Developed an exact match table with millions of entries using DDR, implementing flexible entry space management and accelerating performance with on-chip table cache.

DPDK - DDoS Filter Unit and SDN Switch Multi-Core Performance Optimization (Completed)
Optimized SDN switch performance and developed a method to filter DDoS traffic through IP whitelists and TTL values. The innovation has been patented.

Research Interests 🔬

Network Performance Optimization:
Jiandong Ma's primary research interest is in network performance optimization, where he has contributed significantly to the development of advanced solutions such as high-performance RDMA NICs. His work focuses on enhancing data transmission rates and reducing latency, particularly in complex network environments.

RDMA NIC Development:
One of Jiandong's notable innovations is the development of RDMA NICs that support out-of-order packets. This innovation improves WQE transmission performance in multipath scenarios, demonstrating his expertise in high-speed data transmission technologies.

Packet Reordering and Deduplication Systems:
Jiandong has also worked on developing packet reordering and deduplication systems, which are vital in improving the efficiency and throughput of SD-WAN networks. His solutions have led to reductions in space usage and higher data transmission speeds, addressing critical challenges in wide-area networking.

SDN Switch Performance Optimization:
Another key area of Jiandong’s research involves the optimization of SDN switch performance. He focuses on multi-core performance optimization and DDoS traffic filtering through advanced methods like IP whitelists and TTL values, enhancing network security and reliability.

Flow Control and DDoS Traffic Filtering:
Jiandong is also passionate about designing solutions that improve flow control and network security. His work in DDoS traffic filtering aims to protect networks from malicious attacks while ensuring seamless data flow, making his contributions invaluable to the field of network management and security.

Publications 📚

SSPRD: A Shared-Storage-Based Hardware Packet Reordering and Deduplication System for Multipath Transmission in Wide Area Networks - SCI, accepted

ORNIC: A High-Performance RDMA NIC with Out-of-Order Packet Direct Write Method for Multipath Transmission - SCI, accepted