Ahmad Sedigh Baroughi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Ahmad Sedigh Baroughi has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Akhil Raj Baranwal
School of Engineering Science, Simon Fraser University, Burnaby, British Columbia, Canada
2 shared publications - Hadi Shahriar Shahhoseini
Iran University of Science and Technology, Tehran, Iran
2 shared publications - Nima TaheriNejad
Ruprecht-Karls-Universität Heidelberg, Heidelberg, Germany and TU Wien, Vienna, Austria
2 shared publications - Zhenman Fang
University of Minnesota, USA / Simon Fraser University, Canada
2 shared publications - Jason Cong
Computer Science Department University of California Los Angeles CA USA
1 shared publication - Licheng Guo
University of California Los Angeles, USA
1 shared publication
Latest Publications
- Open Access
- 6 March 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 19 (1) , 1-25
HiSpMM: High Performance High Bandwidth Sparse-Dense Matrix Multiplication on HBM-equipped FPGAs
Sparse Matrix-Dense Matrix Multiplication (SpMM) is a critical operation in scientific computing, machine learning, and graph analytics. However, accelerating SpMM on FPGAs presents major challenges due to irregular memory access patterns and imbalanced workload distribution. In this work, we address a fundamental bottleneck in SpMM acceleration on High Bandwidth Memory (HBM)-equipped FPGAs: workload imbalance among processing elements (PEs). Additionally, we mitigate a scalability barrier present in state-of-the-art designs—namely, the tight coupling between PEs and HBM channels for dense matrix access. Furthermore, we provide an automated design space exploration framework. We propose HiSpMM, a high-performance SpMM accelerator architecture that introduces Dense Row Sharing to mitigate PE under-utilization by distributing heavy-row computations, a decoupled HBM access mechanism to allow independent scaling of PEs and memory bandwidth, and an automation tool that optimizes design parameters according to matrix structure-specific properties and user-defined hardware constraints. Our design achieves a geomean of \(5.81\times\) speedup and \(5.75\times\) energy efficiency improvement for imbalanced matrices compared to state-of-the-art designs, while also maintaining competitive performance for balanced matrices on the AMD/Xilinx U280 HBM FPGA board. Our HiSpMM project will be open sourced in the near future at https://github.com/SFU-HiAccel/HiSpMM .- 6 March 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 19 (1) , 1-25
Open AccessHiSpMM: High Performance High Bandwidth Sparse-Dense Matrix Multiplication on HBM-equipped FPGAs
Sparse Matrix-Dense Matrix Multiplication (SpMM) is a critical operation in scientific computing, machine learning, and graph analytics. However, accelerating SpMM on FPGAs presents major challenges due to irregular memory access patterns and imbalanced workload distribution. In this work, we address a fundamental bottleneck in SpMM acceleration on High Bandwidth Memory (HBM)-equipped FPGAs: workload imbalance among processing elements (PEs). Additionally, we mitigate a scalability barrier present in state-of-the-art designs—namely, the tight coupling between PEs and HBM channels for dense matrix access. Furthermore, we provide an automated design space exploration framework. We propose HiSpMM, a high-performance SpMM accelerator architecture that introduces Dense Row Sharing to mitigate PE under-utilization by distributing heavy-row computations, a decoupled HBM access mechanism to allow independent scaling of PEs and memory bandwidth, and an automation tool that optimizes design parameters according to matrix structure-specific properties and user-defined hardware constraints. Our design achieves a geomean of \(5.81\times\) speedup and \(5.75\times\) energy efficiency improvement for imbalanced matrices compared to state-of-the-art designs, while also maintaining competitive performance for balanced matrices on the AMD/Xilinx U280 HBM FPGA board. Our HiSpMM project will be open sourced in the near future at https://github.com/SFU-HiAccel/HiSpMM . - Open Access
- 30 September 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 17 (3) , 1-31
PASTA: Programming and Automation Support for Scalable Task-Parallel HLS Programs on Modern Multi-Die FPGAs
In recent years, the adoption of FPGAs in datacenters has increased, with a growing number of users choosing High-Level Synthesis (HLS) as their preferred programming method. While HLS simplifies FPGA programming, one notable challenge arises when scaling up designs for modern datacenter FPGAs that comprise multiple dies. The extra delays introduced due to die crossings and routing congestion can significantly degrade the frequency of large designs on these FPGA boards. Due to the gap between HLS design and physical design, it is challenging for HLS programmers to analyze and identify the root causes, and fix their HLS design to achieve better timing closure. Recent efforts have aimed to address these issues by employing coarse-grained floorplanning and pipelining strategies on task-parallel HLS designs where multiple tasks run concurrently and communicate through FIFO stream channels. However, many applications are not streaming friendly and many existing accelerator designs heavily rely on buffer channel based communication between tasks.In this work, we take a step further to support a task-parallel programming model where tasks can communicate via both FIFO stream channels and buffer channels. To achieve this goal, we design and implement the PASTA framework, which takes a large task-parallel HLS design as input and automatically generates a high-frequency FPGA accelerator via HLS and physical design co-optimization. Our framework introduces a latency-insensitive buffer channel design, which supports memory partitioning and ping-pong buffering while remaining compatible with vendor HLS tools. On the frontend, we provide an easy-to-use programming model for utilizing the proposed buffer channel; while on the backend, we implement efficient placement and pipelining strategies for the proposed buffer channel. To validate the effectiveness of our framework, we test it on four widely used Rodinia HLS benchmarks and two real-world accelerator designs and show an average frequency improvement of 25%, with peak improvements of up to 89% on AMD/Xilinx Alveo U280 boards compared to Vitis HLS baselines.
- 30 September 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 17 (3) , 1-31
Open AccessPASTA: Programming and Automation Support for Scalable Task-Parallel HLS Programs on Modern Multi-Die FPGAs
In recent years, the adoption of FPGAs in datacenters has increased, with a growing number of users choosing High-Level Synthesis (HLS) as their preferred programming method. While HLS simplifies FPGA programming, one notable challenge arises when scaling up designs for modern datacenter FPGAs that comprise multiple dies. The extra delays introduced due to die crossings and routing congestion can significantly degrade the frequency of large designs on these FPGA boards. Due to the gap between HLS design and physical design, it is challenging for HLS programmers to analyze and identify the root causes, and fix their HLS design to achieve better timing closure. Recent efforts have aimed to address these issues by employing coarse-grained floorplanning and pipelining strategies on task-parallel HLS designs where multiple tasks run concurrently and communicate through FIFO stream channels. However, many applications are not streaming friendly and many existing accelerator designs heavily rely on buffer channel based communication between tasks.In this work, we take a step further to support a task-parallel programming model where tasks can communicate via both FIFO stream channels and buffer channels. To achieve this goal, we design and implement the PASTA framework, which takes a large task-parallel HLS design as input and automatically generates a high-frequency FPGA accelerator via HLS and physical design co-optimization. Our framework introduces a latency-insensitive buffer channel design, which supports memory partitioning and ping-pong buffering while remaining compatible with vendor HLS tools. On the frontend, we provide an easy-to-use programming model for utilizing the proposed buffer channel; while on the backend, we implement efficient placement and pipelining strategies for the proposed buffer channel. To validate the effectiveness of our framework, we test it on four widely used Rodinia HLS benchmarks and two real-world accelerator designs and show an average frequency improvement of 25%, with peak improvements of up to 89% on AMD/Xilinx Alveo U280 boards compared to Vitis HLS baselines.
- Open Access
- 1 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 71 (5) , 2280-2293
ACE-CNN: Approximate Carry Disregard Multipliers for Energy-Efficient CNN-Based Image Classification
This paper presents the design and development of Signed Carry Disregard Multiplier (SCDM8), a family of signed approximate multipliers tailored for integration into Convolutional Neural Networks (CNNs). Extensive experiments were conducted on popular pre-trained CNN models, including VGG16, VGG19, ResNet101, ResNet152, MobileNetV2, InceptionV3, and ConvNeXt-T to evaluate the trade-off between accuracy and approximation. The results demonstrate that ACE-CNN outperforms other configurations, offering a favorable balance between accuracy and computational efficiency. In our experiments, when applied to VGG16, SCDM8 achieves an average reduction in power consumption of 35% with a marginal decrease in accuracy of only 1.5%. Similarly, when incorporated into ResNet152, SCDM8 yields an energy saving of 42% while sacrificing only 1.8% in accuracy. ACE-CNN provides the first approximate version of ConvNeXt which yields up to 72% energy improvement at the price of less than only 1.3% Top-1 accuracy. These results highlight the suitability of SCDM8 as an approximation method across various CNN models. Our analysis shows that the ACE-CNN outperforms state-of-the-art approaches in accuracy, energy efficiency, and computation precision for image classification tasks in CNNs. Our study investigated the resiliency of CNN models to approximate multipliers, revealing that ResNet101 demonstrated the highest resiliency with an average difference in the accuracy of 0.97%, whereas LeNet5 Inspired-CNN exhibited the lowest resiliency with an average difference of 2.92%. These findings aid in selecting energy-efficient approximate multipliers for CNN-based systems, and contribute to the development of energy-efficient deep learning systems by offering an effective approximation technique for multipliers in CNNs. The proposed SCDM8 family of approximate multipliers opens new avenues for efficient deep learning applications, enabling significant energy savings with virtually no loss in accuracy.
- 1 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 71 (5) , 2280-2293
Open AccessACE-CNN: Approximate Carry Disregard Multipliers for Energy-Efficient CNN-Based Image Classification
This paper presents the design and development of Signed Carry Disregard Multiplier (SCDM8), a family of signed approximate multipliers tailored for integration into Convolutional Neural Networks (CNNs). Extensive experiments were conducted on popular pre-trained CNN models, including VGG16, VGG19, ResNet101, ResNet152, MobileNetV2, InceptionV3, and ConvNeXt-T to evaluate the trade-off between accuracy and approximation. The results demonstrate that ACE-CNN outperforms other configurations, offering a favorable balance between accuracy and computational efficiency. In our experiments, when applied to VGG16, SCDM8 achieves an average reduction in power consumption of 35% with a marginal decrease in accuracy of only 1.5%. Similarly, when incorporated into ResNet152, SCDM8 yields an energy saving of 42% while sacrificing only 1.8% in accuracy. ACE-CNN provides the first approximate version of ConvNeXt which yields up to 72% energy improvement at the price of less than only 1.3% Top-1 accuracy. These results highlight the suitability of SCDM8 as an approximation method across various CNN models. Our analysis shows that the ACE-CNN outperforms state-of-the-art approaches in accuracy, energy efficiency, and computation precision for image classification tasks in CNNs. Our study investigated the resiliency of CNN models to approximate multipliers, revealing that ResNet101 demonstrated the highest resiliency with an average difference in the accuracy of 0.97%, whereas LeNet5 Inspired-CNN exhibited the lowest resiliency with an average difference of 2.92%. These findings aid in selecting energy-efficient approximate multipliers for CNN-based systems, and contribute to the development of energy-efficient deep learning systems by offering an effective approximation technique for multipliers in CNNs. The proposed SCDM8 family of approximate multipliers opens new avenues for efficient deep learning applications, enabling significant energy savings with virtually no loss in accuracy.
- 31 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-7
Approximation-aware Task Partitioning on an Approximate-Exact MPSoC (AxE)
As the demand for increased performance and reduced energy consumption continues to grow, Quality of Service (QoS) adjustment approaches offer an effective way to tackle those demands. One such method, approximation, has gained popularity in recent years, facilitating faster executions as well as a smaller power consumption by providing an approximated result. The areas in which these trade-offs are acceptable are numerous, but hardware-based solutions are usually domain-specific and expensive to integrate. To tackle this issue, we take a different approach, in which approximate hardware can be used (or not) in a general purpose environment and via software decisions. That is, a Multi-Processor System-on-Chip (MPSoC) that contains Central Processing Units (CPUs) that offer approximate calculations alongside the ones that offer exact calculations. However, current task partitioning algorithms do not consider the specific capabilities or requirements of such a MPSoC. This paper introduces approximation-aware partitioning algorithms using different strategies and compares the results to the State-of-the-Art (SoA). Additionally, the resulted task partitions are executed to gauge their quality compared to the SoA. Experimental results show, that the usage of an approximate CPU and approximation-aware task partitioning leads to an increased partition success rate of 21.5%. Furthermore, the execution, i.e., scheduling of the partitioned tasks until energy starvation, achieves a 3.4% extended run-time.
- 31 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-7
Approximation-aware Task Partitioning on an Approximate-Exact MPSoC (AxE)
As the demand for increased performance and reduced energy consumption continues to grow, Quality of Service (QoS) adjustment approaches offer an effective way to tackle those demands. One such method, approximation, has gained popularity in recent years, facilitating faster executions as well as a smaller power consumption by providing an approximated result. The areas in which these trade-offs are acceptable are numerous, but hardware-based solutions are usually domain-specific and expensive to integrate. To tackle this issue, we take a different approach, in which approximate hardware can be used (or not) in a general purpose environment and via software decisions. That is, a Multi-Processor System-on-Chip (MPSoC) that contains Central Processing Units (CPUs) that offer approximate calculations alongside the ones that offer exact calculations. However, current task partitioning algorithms do not consider the specific capabilities or requirements of such a MPSoC. This paper introduces approximation-aware partitioning algorithms using different strategies and compares the results to the State-of-the-Art (SoA). Additionally, the resulted task partitions are executed to gauge their quality compared to the SoA. Experimental results show, that the usage of an approximate CPU and approximation-aware task partitioning leads to an increased partition success rate of 21.5%. Furthermore, the execution, i.e., scheduling of the partitioned tasks until energy starvation, achieves a 3.4% extended run-time.