Muhammad Ali
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Muhammad Ali has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
- Diana Göhringer
Chair of Adaptive Dynamic System, Technische Universitt Dresden, Dresden, Germany
7 shared publications - Alireza Syavashi2 shared publications
- Christian Herglotz
Chair of Computer Engineering, Brandenburgische Technische Universitt Cottbus-Senftenberg (BTU), Cottbus, Germany
2 shared publications - Ensieh Aliagha
Chair of Adaptive Dynamic System, Technische Universitt Dresden, Dresden, Germany
2 shared publications - Hector Gerardo Munoz Hernandez
Brandenburg University of Technology Cottbus-Senftenberg
2 shared publications - Keyvan Shahin
Chair of Computer Engineering, Brandenburgische-Technische Universitt CottbusSenftenberg (BTU), Cottbus, Germany
2 shared publications
Latest Publications
- Open Access
- 27 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 34545-34563
Application-Specific Instruction-Set Processors (ASIPs) for Deep Neural Networks: A Survey
The demand for artificial intelligence applications is rising in every field. And with this demand, machine learning algorithms and techniques are becoming more complicated and compute-intensive. For embedded devices, these compute-intensive and high memory-constrained algorithms are becoming a challenge. Dedicated hardware accelerators are generally the leading solution for embedded systems. This is because of their high performance, low area, and power footprint. Although hardware accelerators are a great solution for machine learning, they lack flexibility and programmability. Another alternative is General- Purpose Processors (GPPs), which provide better flexibility and programmability as compared with hardware accelerators, however, they lack performance, area, and power. Between these two plausible solutions, there is a big gap in terms of different metrics, which can be bridged by Application-Specific Instruction-set Processors (ASIPs). ASIPs are specialized processor systems with an architecture that is tailor-made for a certain application. ASIPs provide better performance as compared with general-purpose processors and have more flexibility and programmability as compared with hardware accelerators. The main goal of an ASIP design is to maximize the performance-to-power ratio for a specific application. This work provides an in-depth survey of different ASIP designs with a focus on Deep Neural Networks (DNNs). For a generic comparison, the proposed related works of ASIP design are classified based on their microarchitecture approach and optimization adopted. The strong and weak points are pointed out in different optimization and microarchitectures, and future trends for ASIP design for fast machine learning are identified.
- 27 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 34545-34563
Open AccessApplication-Specific Instruction-Set Processors (ASIPs) for Deep Neural Networks: A Survey
The demand for artificial intelligence applications is rising in every field. And with this demand, machine learning algorithms and techniques are becoming more complicated and compute-intensive. For embedded devices, these compute-intensive and high memory-constrained algorithms are becoming a challenge. Dedicated hardware accelerators are generally the leading solution for embedded systems. This is because of their high performance, low area, and power footprint. Although hardware accelerators are a great solution for machine learning, they lack flexibility and programmability. Another alternative is General- Purpose Processors (GPPs), which provide better flexibility and programmability as compared with hardware accelerators, however, they lack performance, area, and power. Between these two plausible solutions, there is a big gap in terms of different metrics, which can be bridged by Application-Specific Instruction-set Processors (ASIPs). ASIPs are specialized processor systems with an architecture that is tailor-made for a certain application. ASIPs provide better performance as compared with general-purpose processors and have more flexibility and programmability as compared with hardware accelerators. The main goal of an ASIP design is to maximize the performance-to-power ratio for a specific application. This work provides an in-depth survey of different ASIP designs with a focus on Deep Neural Networks (DNNs). For a generic comparison, the proposed related works of ASIP design are classified based on their microarchitecture approach and optimization adopted. The strong and weak points are pointed out in different optimization and microarchitectures, and future trends for ASIP design for fast machine learning are identified.
- Open Access
- 23 February 2026
- Published by Elsevier in Journal of Systems Architecture
- Vol. 175
Integrating an open-source soft-GPU overlay with RISC-V control and high-bandwidth memory
- 23 February 2026
- Published by Elsevier in Journal of Systems Architecture
- Vol. 175
Open AccessIntegrating an open-source soft-GPU overlay with RISC-V control and high-bandwidth memory
- 23 October 2025
- Published by Springer Nature
- p. 94-108
Towards Complete Open-Source Environments: FPGA-Based GPU Overlay Controlled by RISC-V
- 23 October 2025
- Published by Springer Nature
- p. 94-108
Towards Complete Open-Source Environments: FPGA-Based GPU Overlay Controlled by RISC-V
- Open Access
- 19 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 146603-146616
P-CORE: Exploring RISC-V Packed-SIMD Extension for CNNs
In today’s technological landscape, embedded and IoT devices face escalating demands for performance and power efficiency in inference tasks employing Convolution Neural Networks (CNNs). Various methods have emerged to improve performance, including quantization and the adoption of fixed-point formats instead of floating-point formats. RISC-V Instruction Set Architecture (ISA) being open-source and with a modular specification poses as a suitable platform for targeting such applications for embedded processing. The RISC-V’s standard P-extension represents a significant advancement in integer-based computation, offering new ways for achieving higher speedups through integer-packed SIMD instructions. This work presents the first RISC-V processor based on RV32IM base ISA with an optional support of P-extension Version 0.9.11-draft-20211209 called the P-CORE. The P-CORE implements P-extension ISA for RV32 specification. The implemented processor is evaluated for Xilinx Zynq Ultrascale+ FPGA using different test case functions; matrix multiplication, convolution, max-pooling, and fully connected. A CNN used for detecting handwritten digits called Lenet-5 is also used as a test case to evaluate the P-CORE. The evaluation also presents Design Space Exploration (DSE) for P-extension by exploiting the level of data-level-parallelism; SIMD8 (8-bit (4-way)) and SIMD16 (16-bit (2-way)). Our investigation demonstrates that with the P-extension activated, CNN algorithm computations can achieve speedups of up to 17x for max-pooling, 7x for matrix multiplication, and 4.8x for fully connected networks, accompanied by similar enhancements in power efficiency. The processor achieved a maximum performance of 335.1 MIOPS and 490.6 MIOPS/W efficiency for matrix multiplication using -O3 compiler optimization.
- 19 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 146603-146616
Open AccessP-CORE: Exploring RISC-V Packed-SIMD Extension for CNNs
In today’s technological landscape, embedded and IoT devices face escalating demands for performance and power efficiency in inference tasks employing Convolution Neural Networks (CNNs). Various methods have emerged to improve performance, including quantization and the adoption of fixed-point formats instead of floating-point formats. RISC-V Instruction Set Architecture (ISA) being open-source and with a modular specification poses as a suitable platform for targeting such applications for embedded processing. The RISC-V’s standard P-extension represents a significant advancement in integer-based computation, offering new ways for achieving higher speedups through integer-packed SIMD instructions. This work presents the first RISC-V processor based on RV32IM base ISA with an optional support of P-extension Version 0.9.11-draft-20211209 called the P-CORE. The P-CORE implements P-extension ISA for RV32 specification. The implemented processor is evaluated for Xilinx Zynq Ultrascale+ FPGA using different test case functions; matrix multiplication, convolution, max-pooling, and fully connected. A CNN used for detecting handwritten digits called Lenet-5 is also used as a test case to evaluate the P-CORE. The evaluation also presents Design Space Exploration (DSE) for P-extension by exploiting the level of data-level-parallelism; SIMD8 (8-bit (4-way)) and SIMD16 (16-bit (2-way)). Our investigation demonstrates that with the P-extension activated, CNN algorithm computations can achieve speedups of up to 17x for max-pooling, 7x for matrix multiplication, and 4.8x for fully connected networks, accompanied by similar enhancements in power efficiency. The processor achieved a maximum performance of 335.1 MIOPS and 490.6 MIOPS/W efficiency for matrix multiplication using -O3 compiler optimization.
- 28 January 2025
- Published by Springer Nature
- p. 59-71
RV-VP${^{2}}$: Unlocking the Potential of RISC-V Packed-SIMD for Embedded Processing- 28 January 2025
- Published by Springer Nature
- p. 59-71
RV-VP${^{2}}$: Unlocking the Potential of RISC-V Packed-SIMD for Embedded Processing- 7 November 2023
- Published by Springer Nature
- p. 269-282
Unlocking the Potential of RISC-V Heterogeneous MPSoC: A PANACA-Based Approach to Simulation and Modeling
- 7 November 2023
- Published by Springer Nature
- p. 269-282
Unlocking the Potential of RISC-V Heterogeneous MPSoC: A PANACA-Based Approach to Simulation and Modeling
- Open Access
- 15 March 2021
- Published by Springer Nature in Journal of Signal Processing Systems
- Vol. 93 (5) , 513-529
A Parametrizable High-Level Synthesis Library for Accelerating Neural Networks on FPGAs
In recent years, Convolutional Neural Network CNN have been incorporated in a large number of applications, including multimedia retrieval and image classification. However, CNN based algorithms are computationally and resource intensive and therefore difficult to be used in embedded systems. FPGA based accelerators are becoming more and more popular in research and industry due to their flexibility and energy efficiency. However, the available resources and the size of the on-chip memory can limit the performance of the FPGA accelerator for CNN. This work proposes an High-Level Synthesis HLS library for CNN algorithms. It contains seven different streaming-capable CNN (plus two conversion) functions for creating large neural networks with deep pipelines. The different functions have many parameter settings (e.g. for resolution, feature maps, data types, kernel size, parallelilization, accuracy, etc.), which also enable compile-time optimizations. Our functions are integrated into the HiFlipVX library, which is an open source HLS FPGA library for image processing and object detection. This offers the possibility to implement different types of computer vision applications with one library. Due to the various configuration and parallelization possibilities of the library functions, it is possible to implement a high-performance, scalable and resource-efficient system, as our evaluation of the MobileNets algorithm shows.
- 15 March 2021
- Published by Springer Nature in Journal of Signal Processing Systems
- Vol. 93 (5) , 513-529
Open AccessA Parametrizable High-Level Synthesis Library for Accelerating Neural Networks on FPGAs
In recent years, Convolutional Neural Network CNN have been incorporated in a large number of applications, including multimedia retrieval and image classification. However, CNN based algorithms are computationally and resource intensive and therefore difficult to be used in embedded systems. FPGA based accelerators are becoming more and more popular in research and industry due to their flexibility and energy efficiency. However, the available resources and the size of the on-chip memory can limit the performance of the FPGA accelerator for CNN. This work proposes an High-Level Synthesis HLS library for CNN algorithms. It contains seven different streaming-capable CNN (plus two conversion) functions for creating large neural networks with deep pipelines. The different functions have many parameter settings (e.g. for resolution, feature maps, data types, kernel size, parallelilization, accuracy, etc.), which also enable compile-time optimizations. Our functions are integrated into the HiFlipVX library, which is an open source HLS FPGA library for image processing and object detection. This offers the possibility to implement different types of computer vision applications with one library. Due to the various configuration and parallelization possibilities of the library functions, it is possible to implement a high-performance, scalable and resource-efficient system, as our evaluation of the MobileNets algorithm shows.
- 25 March 2020
- Published by Springer Nature
- p. 193-207
RISC-V Based MPSoC Design Exploration for FPGAs: Area, Power and Performance
- 25 March 2020
- Published by Springer Nature
- p. 193-207
RISC-V Based MPSoC Design Exploration for FPGAs: Area, Power and Performance
- 1 January 2016
- Published by INSTICC
- p. 449-457
A Holistic Method to Recognize Characters in Natural Scenes
- 1 January 2016
- Published by INSTICC
- p. 449-457
A Holistic Method to Recognize Characters in Natural Scenes
- 1 January 2016
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 373-380
QoS and security in VOIP networks through admission control mechanism
With the developing understanding of Information Security and digital assets, IT technology has put on tremendous importance of network admission control (NAC). In NAC architecture, admission decisions and resource reservations are taken at edge devices, rather than resources or individual routers within the network. The NAC architecture enables resilient resource reservation, maintaining reservations even after failures and intra-domain rerouting. Admission Control Networks destiny is based on IP networks through its Security and Quality of Service (QoS) demands for real time multimedia application via advance resource reservation techniques. To achieve Security & QoS demands, in real time performance networks, admission control algorithm decides whether the new traffic flow can be admitted to the network or not. Secure allocation of Peer for multimedia traffic flows with required performance is a great challenge in resource reservation schemes. In this paper, we have proposed our model for VoIP networks in order to achieve security services along with QoS, where admission control decisions are taken place at edge routers. We have analyzed and argued that the measurement based admission control should be done at edge routers which employs on-demand probing parallel from both edge routers to secure the source and destination nodes respectively. In order to achieve Security and QoS for a new call, we choose various probe packet sizes for voice and video calls respectively. Similarly a technique is adopted to attain a security allocation approach for selecting an admission control threshold by proposing our admission control algorithm. All results are tested on NS2 based simulation to evalualate the network performance of edge router based upon network admission control in VoIP traffic.
- 1 January 2016
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 373-380
QoS and security in VOIP networks through admission control mechanism
With the developing understanding of Information Security and digital assets, IT technology has put on tremendous importance of network admission control (NAC). In NAC architecture, admission decisions and resource reservations are taken at edge devices, rather than resources or individual routers within the network. The NAC architecture enables resilient resource reservation, maintaining reservations even after failures and intra-domain rerouting. Admission Control Networks destiny is based on IP networks through its Security and Quality of Service (QoS) demands for real time multimedia application via advance resource reservation techniques. To achieve Security & QoS demands, in real time performance networks, admission control algorithm decides whether the new traffic flow can be admitted to the network or not. Secure allocation of Peer for multimedia traffic flows with required performance is a great challenge in resource reservation schemes. In this paper, we have proposed our model for VoIP networks in order to achieve security services along with QoS, where admission control decisions are taken place at edge routers. We have analyzed and argued that the measurement based admission control should be done at edge routers which employs on-demand probing parallel from both edge routers to secure the source and destination nodes respectively. In order to achieve Security and QoS for a new call, we choose various probe packet sizes for voice and video calls respectively. Similarly a technique is adopted to attain a security allocation approach for selecting an admission control threshold by proposing our admission control algorithm. All results are tested on NS2 based simulation to evalualate the network performance of edge router based upon network admission control in VoIP traffic.