Boming Huang

Boming Huang

Auto-generated profile, based on publicly available information

Boming Huang

Microelectronics Research Institute, Hangzhou Dianzi University, Hangzhou, China

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

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

Representative Publications

View All Publications
4 October 2021 in IEEE Transactions on Circuits and Systems I: Regular Papers
Verified
Journal Article
Closed access
BHBoming Huang
Boming Huang
YHYuxiang Huan
Yuxiang Huan
HCHaoming Chu
Haoming Chu
JXJiawei Xu
Jiawei Xu
LLLizheng Liu
Lizheng Liu
ZZZHUO ZHUO
ZHUO ZHUO
It remains challenging for a Convolutional Neural Network (CNN) accelerator to maintain high hardware utilization and low processing latency with restricted on-chip memory. This paper presents an In-Execution Configuration Accelerator (IECA) that realizes an efficient control scheme, exploring architectural data reuse, unified in-execution controlling, and pipelined latency hiding to minimize configuration overhead out of the computation scope. The proposed IECA achieves row-wise convolution with tiny distributed buffers and reduces the size of total on-chip memory by removing 40% of redundant memory storage with shared delay chains. By exploiting a reconfigurable Sequence Mapping Table (SMT) and Finite State Machine (FSM) control, the chip realizes cycle-accurate Processing Element (PE) control, automatic loop tiling and latency hiding without extra time slots for pre-configuration. Evaluated on AlexNet and VGG-16, the IECA retains over 97.3% PE utilization and over 95.6% memory access time hiding on average. The chip is designed and fabricated in a UMC 55-nm process running at a frequency of 250 MHz and achieves an area efficiency of 30.55 GOPS/mm 2 and 0.244 GOPS/KGE (kilo-gate-equivalent), which makes an over $2.0\times $ and $2.1\times $ improvement, respectively, compared with that of previous related works. Implementation of the IEC control scheme uses only a 0.55% area of the 2.75 mm 2 core.

2 August 2024 in IEEE Journal on Emerging and Selected Topics in Circuits and Systems
Verified
Journal Article
Closed access
HJHuidong Ji
Huidong Ji
CDChen Ding
Chen Ding
BHBoming Huang
Boming Huang
YHYuxiang Huan
Yuxiang Huan
LIRONG ZHENGLIRONG ZHENG
LIRONG ZHENG
ZZZHUO ZHUO
ZHUO ZHUO
Exploding development of convolutional neural network (CNN) benefits greatly from the hardware-based acceleration to maintain low latency and high utilization of resources. To enhance the processing efficiency of CNN algorithms, Field Programming Gate Array (FPGA)-based accelerators are designed with increased hardware resources to achieve high parallelism and throughput. However, there exist bottlenecks when more processing elements (PEs) in the form of PE clusters are introduced, including (1) the under-utilization of FPGA’s fixed hardware resources, which leads to the effective and peak performance mismatch, and (2) the limited clock frequency caused by the sophisticated routing and complex placement. In this paper, a 2-level hierarchical Network-on-Chip (NoC)-based CNN accelerator is proposed. In the upper level, a mesh-based NoC that interconnects multiple PE clusters is introduced. Such a design not only provides increased flexibility to balance different data communication models for better PE utilization and energy efficiency but also enables globally asynchronous, locally synchronous (GALS) architecture for better timing closure. At the lower level, local PEs are organized into a 3D-tiled PE cluster aiming to maximize the data reuse exploiting inherent dataflow of the convolution networks. Implementation and experiments on Xilinx ZU9EG FPGA for 4 benchmark CNN models: ResNet50, ResNet34, VGG16, and Darknet19 show that our work operates at a frequency of 300 MHz and delivers an effective throughput of 0.998 TOPS, 1.022 TOPS, 1.024 TOPS, and 1.026 TOPS. This result corresponds to 92.85%, 95.1%, 95.25%, and 95.46% PE utilization. Compared with the related FPGA-based designs, our work improves the resource efficiency of DSP by 5.36×, 1.62×, 1.96×, and 5.83×, respectively.

20 September 2024 in IEEE Embedded Systems Letters
Verified
Journal Article
Closed access
JLJianjun Luo
Jianjun Luo
YSYifan Shen
Yifan Shen
BHBoming Huang
Boming Huang
MEMichael Etzkorn
Michael Etzkorn
HCHongqiang Chen
Hongqiang Chen
CYCong Yu
Cong Yu
Low Density Parity Check Code (LDPC) is the most popular Error Correction Code (ECC) for current NAND flash memory controllers. Next generation flash memory, such as Triple-Level Cell (TLC) and Quad-Level Cell (QLC) with a higher bit error rate, bring up the demand for superior ECC algorithms providing excellent performance and acceptable hardware overhead. This paper proposes Turbo Product Code (TPC) as an alternative to LDPC by exploring embedding of a TPC engine into a Solid-State Drive (SSD) controller architecture. The implementation of this TPC engine uses two-dimensional error coding and applies the Bose-Chaudhuri-Hocquenghem (BCH) code in each dimension. The algorithm is further improved by flipping the dedicated bit matrix when the basic TPC decoding algorithm terminates with uncorrectable conditions. This Bit-Flipping Enhanced TPC (BFE-TPC) module is finally integrated into a Non-Volatile Memory Express (NVMe) SSD controller driving eight flash memory channels. This BFE-TPC module illustrates the potential of 2D-TPCs as a replacement for LDPCs due to its high throughput and low hardware overhead.

13 April 2022 in IEEE Transactions on Circuits and Systems II: Express Briefs
Verified
Journal Article
Closed access
BHBoming Huang
Boming Huang
YHYuxiang Huan
Yuxiang Huan
HJHao Jia
Hao Jia
CDChen Ding
Chen Ding
YYYulong Yan
Yulong Yan
ZZZHUO ZHUO
ZHUO ZHUO
Latency of a trading system is crucial for gaining profitability in the financial market. Traditional software-based trading system suffers from high latency in market data parsing, network communication and operating system scheduling. The latency is even worse if algorithm processing is used for market data analysis, especially Deep Neural Networks (DNN) are involved. This brief presents a hardware-based trading system which achieves all key functionalities including network communication, financial protocol parsing and DNN processing in one FPGA card. Running at 300 MHz, the proposed hardware design realizes over $98\times $ acceleration than the software-based system regarding the overall latency.