Yu Huang

Yu Huang

Auto-generated profile, based on publicly available information

Yu Huang

Huazhong University of Science and Technology, Wuhan City, Hubei, China

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
17 June 2023
Verified
Conference Paper
Open access
DCDan Chen
Dan Chen
HHHaiheng He
Haiheng He
HJHai Jin
Hai Jin
LZLong Zheng
Long Zheng
YHYu Huang
Yu Huang
XLXiaoFei Liao
XiaoFei Liao
Heterogeneous graph neural networks (HGNNs) based on metapath exhibit powerful capturing of rich structural and semantic information in the heterogeneous graph. HGNNs are highly memory-bound and thus can be accelerated by near-memory processing. However, they also suffer from significant memory footprint (due to storing metapath instances as intermediate data) and severe redundant computation (when vertex features are aggregated among metapath instances). To address these issues, this paper proposes MetaNMP, the first DIMM-based near-memory processing HGNNs accelerator with reduced memory footprint and high performance. Specifically, we first propose a cartesian-like product paradigm to generate all metapath instances on the fly for heterogeneous graphs. In this way, metapath instances no longer need to be stored as intermediate data, avoiding significant memory consumption. We then design a data flow for aggregating vertex features on metapath instances, which aggregates vertex features along the direction of the metapath instances dispersed from the starting vertex to exploit shareable aggregation computations, eliminating most of the redundant computations. Finally, we integrate specialized hardware units in DIMM to accelerate HGNNs with near-memory processing, and introduce a broadcast mechanism for edge data and vertex features to mitigate the inter-DIMM communication. Our evaluation shows that MetaNMP achieves the memory space reduction of 51.9% on average and the performance improvement by 415.18× compared to NVIDIA Tesla V100 GPU.

17 June 2023
Verified
Conference Paper
Open access
HLHaifeng Liu
Haifeng Liu
LZLong Zheng
Long Zheng
YHYu Huang
Yu Huang
CLChaoqiang Liu
Chaoqiang Liu
XYXiangyu Ye
Xiangyu Ye
JXJingling Xue
Jingling Xue
The memory-intensive embedding layers of the personalized recommendation systems are the performance bottleneck as they demand large memory bandwidth and exhibit irregular and sparse memory access patterns. Recent studies propose near memory processing (NMP) to accelerate memory-bound embedding operations. However, due to the load imbalance caused by the skewed access frequency of the embedding data, existing NMP solutions that exploit fine-grained memory parallelism fail to translate the increasingly massive internal bandwidth to performance improvements, leading to resource underutilization and hardware overhead. We propose an efficient yet practical fine-grained NMP accelerator for embedding operations. We architect ReCross, a cross-level NMP architecture that exploits rank, bank-group, and subarray-level memory parallelism in a unified DIMM-based memory system by supporting rank, bank-group, and bank-level NMP to accommodate various bandwidth requirements of embedding data. In addition, we present a novel embedding partitioning technique to quantify the bandwidth requirements of embedding tables and allocate them to appropriate NMP levels. ReCross innovatively collaborates the data and architecture characteristics for the NMP embedding layer acceleration, achieving high resource utilization and performance. Our evaluation shows that ReCross outperforms a state-of-the-art bank-group level NMP solution, TRiM-G, by 2.5× with nearly the same area overhead and bank-level NMP solution, TRiM-B, by 1.8× with an area overhead reduction of 4×.

15 March 2023 in IEEE Transactions on Computers
Verified
Journal Article
Open access
HJHai Jin
Hai Jin
DCDan Chen
Dan Chen
LZLong Zheng
Long Zheng
YHYu Huang
Yu Huang
PYPengcheng Yao
Pengcheng Yao
WJWenbin Jiang
Wenbin Jiang
Graph convolutional networks (GCNs) are promising to enable machine learning on graph data. GCNs show potential vertex-level and intra-vertex parallelism for GPU acceleration, but their irregular memory accesses arising in aggregation operations and the inherent sparsity for vertex features of graphs cause inefficiencies on the GPU. In this paper, we present gPIM, which aims to accelerate GCNs inference through a processing-in-memory (PIM) enabled architecture. gPIM is expected to perform compute-intensive combination on the GPU while aggregation and memory-bound combination are offloaded to the PIM-featured hybrid memory cubes (HMCs). To maximize the efficiency of such GPU-HMC architecture, gPIM is novel with two key designs: 1) A GCN-induced graph partitioning that minimizes communication overheads between cubes, 2) A programmer-transparent performance estimation mechanism that predicts the performance bound of operations accurately for workload offloading. Experimental results show that gPIM significantly outperforms Intel Xeon E5-2680v3 CPU (8,979.52×), NVIDIA Tesla V100 GPU (96.01×), and a state-of-the-art GCN accelerator AWB-GCN (4.18×).

20 October 2022 in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
Verified
Journal Article
Closed access
YHYu Huang
Yu Huang
LZLong Zheng
Long Zheng
PYPengcheng Yao
Pengcheng Yao
QWQinggang Wang
Qinggang Wang
HLHaifeng Liu
Haifeng Liu
JXJingling Xue
Jingling Xue
Dynamic graph convolutional networks (DGCNs) have emerged as an effective approach to analyzing graph data that is constantly changing. The typical DGCNs incorporate not only graph convolutional networks (GCNs) to extract the structural information but also with recurrent neural networks (RNNs) to capture the temporal information from evolving graph data. These two alternative execution kernels of DGCNs impose unique architecture challenges for both types of kernels to be implemented efficiently. The presence of complex execution patterns of DGCNs renders existing architectures unsuitable. In this article, we present the first DGCN accelerator with an integrated architecture, named ReaDy, to accelerate DGCNs based on emerging PIM-featured ReRAM architectures. ReaDy is novel with an integrated architecture that enables running the GCN and RNN kernels of DGCNs simultaneously. Specifically, ReaDy is equipped with a redundancy-free scheduling mechanism to alleviate intrinsic dynamic irregularity for the GCN kernel, improving hardware utilization. In addition, ReaDy also includes a locality-aware dataflow strategy to exploit the inherent intervertex data locality for the RNN kernel, reducing superfluous data accesses to vertices and weight parameters. In a holistic view, ReaDy further enhances the entire system via an interkernel pipeline to reduce the off-chip accesses of intermediate results, boosting the overall efficiency of DGCNs significantly. Compared to the state-of-the-art software framework, PyGT, running on Intel Xeon E5-2680v4 CPU and NVIDIA Ampere A100 GPU, ReaDy achieves the average speedups of $955\times $ and $27.33\times $ , and the average energy savings of 1 $093\times $ and $80.21\times $ , respectively. In addition, ReaDy outperforms ReFlip-ERA, which is obtained by combining a state-of-the-art GCN accelerator ReFlip and RNN accelerator ERA-LSTM, by an average speedup of $8.30\times $ and an average energy saving of $7.29\times $ .

12 February 2023
Verified
Conference Paper
Open access
CLChaoqiang Liu
Chaoqiang Liu
HLHaifeng Liu
Haifeng Liu
LZLong Zheng
Long Zheng
YHYu Huang
Yu Huang
XYXiangyu Ye
Xiangyu Ye
HJHai Jin
Hai Jin

26 December 2023 in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
Verified
Journal Article
Open access
HLHuize Li
Huize Li
HJHai Jin
Hai Jin
LZLong Zheng
Long Zheng
XLXiaoFei Liao
XiaoFei Liao
YHYu Huang
Yu Huang
CGChuangyi Gui
Chuangyi Gui
The attention-based neural network attracts great interest due to its excellent accuracy enhancement. However, the attention mechanism requires huge computational efforts to process unnecessary calculations, significantly limiting the system’s performance. To reduce the unnecessary calculations, researchers propose sparse attention to convert some dense-dense matrices multiplication (DDMM) operations to sampled dense-dense matrix multiplication (SDDMM) and sparse matrix multiplication (SpMM) operations. However, current sparse attention solutions introduce massive off-chip random memory access since the sparse attention matrix is generally unstructured. We propose CPSAA, a novel crossbar-based processing-in-memory (PIM)-featured sparse attention accelerator to eliminate off-chip data transmissions. First, we present a novel attention calculation mode to balance the crossbar writing and crossbar processing latency. Second, we design a novel PIM-based sparsity pruning architecture to eliminate the pruning phase’s off-chip data transfers. Finally, we present novel crossbar-based SDDMM and SpMM methods to process unstructured sparse attention matrices by coupling two types of crossbar arrays. Experimental results show that CPSAA has an average of 89.6×, 32.2×, 17.8×, 3.39×, and 3.84× performance improvement and 755.6×, 55.3×, 21.3×, 5.7×, and 4.9× energy-saving when compare with GPU, FPGA, SANGER, ReBERT, and ReTransformer.