Heng Liao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Heng Liao has a total of 38 co-authors. Here are the top co-authors by number of shared publications:
Bill McColl
HUAWEI, Shenzhen, China
1 shared publication- Bingyang Liu
HUAWEI, Shenzhen, China
1 shared publication Buyun Wang
HUAWEI, Shenzhen, China
1 shared publicationChuanning Cheng
HUAWEI, Shenzhen, China
1 shared publicationChungang Li
HUAWEI, Shenzhen, China
1 shared publicationCunle Qian
HUAWEI, Shenzhen, China
1 shared publication
Latest Publications
- Open Access
- 17 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 170-183
RICH Prefetcher: Storing Rich Information in Memory to Trade Capacity and Bandwidth for Latency Hiding
- 17 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 170-183
Open AccessRICH Prefetcher: Storing Rich Information in Memory to Trade Capacity and Bandwidth for Latency Hiding
- Open Access
- 4 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Micro
- Vol. 45 (5) , 20-29
UB-Mesh: A Hierarchically Localized nD-FullMesh Data Center Network Architecture
The scaling of Large-scale Language Models (LLMs) demands unprecedented computational power and bandwidth. We present UB-Mesh, an innovative AI datacenter network architecture that enhances scalability, performance, and cost-efficiency through a hierarchical nD-FullMesh topology. Unlike traditional symmetrical designs, UB-Meshoptimizes LLM training by prioritizing localized data movement and minimizing switch usage. The architecture features UB-Mesh-Pod, a physical implementation of 4D-FullMesh using custom hardware including NPUs, CPUs, Low/High-Radix Switches (LRS/HRS), and NICs, interconnected via our Unified Bus (UB) technology for dynamic resource allocation. For network optimization, we introduce All-Path-Routing (APR) to efficiently manage data traffic. Combined with topology-aware performance tuning and robust reliability mechanisms like 64 + 1 backup, UB-Meshachieves 2.04× better cost-efficiency and 7.2% higher availability than Clos networks. These innovations address the critical challenges of building practical, high-performance AI infrastructure at scale.
- 4 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Micro
- Vol. 45 (5) , 20-29
Open AccessUB-Mesh: A Hierarchically Localized nD-FullMesh Data Center Network Architecture
The scaling of Large-scale Language Models (LLMs) demands unprecedented computational power and bandwidth. We present UB-Mesh, an innovative AI datacenter network architecture that enhances scalability, performance, and cost-efficiency through a hierarchical nD-FullMesh topology. Unlike traditional symmetrical designs, UB-Meshoptimizes LLM training by prioritizing localized data movement and minimizing switch usage. The architecture features UB-Mesh-Pod, a physical implementation of 4D-FullMesh using custom hardware including NPUs, CPUs, Low/High-Radix Switches (LRS/HRS), and NICs, interconnected via our Unified Bus (UB) technology for dynamic resource allocation. For network optimization, we introduce All-Path-Routing (APR) to efficiently manage data traffic. Combined with topology-aware performance tuning and robust reliability mechanisms like 64 + 1 backup, UB-Meshachieves 2.04× better cost-efficiency and 7.2% higher availability than Clos networks. These innovations address the critical challenges of building practical, high-performance AI infrastructure at scale.