Abdallah Cheikh
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
Abdallah Cheikh has a total of 38 co-authors. Here are the top co-authors by number of shared publications:
- Mauro Olivieri
Department of Information Engineering, Electronics and Telecommunications, Sapienza University of Rome, Rome, Italy
20 shared publications - Marcello Barbirotta
Department of Information Engineering, Electronics and Telecommunications, Sapienza University of Rome, Rome, Italy
17 shared publications - Antonio Mastrandrea
Department of Information Engineering, Electronics and Telecommunications (DIET), Sapienza University of Rome, 00184 Rome, Italy
16 shared publications - Francesco Menichelli
Department of Information Engineering, Electronics and Telecommunications, Sapienza University of Rome, Rome, Italy
16 shared publications - Marco Angioli
Sapienza University of Rome
11 shared publications - Saeid Jamili
University of Rome “La Sapienza
6 shared publications
Latest Publications
- 10 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 75 (2) , 653-664
Configurable Hardware Acceleration for Hyperdimensional Computing Extension on RISC-V
Hyperdimensional Computing (HDC) is a neuro-inspired computational model that represents and manipulates information using high-dimensional distributed representations that are combined and compared using simple and highly parallel vector operations. In this work, we present a highly flexible hardware acceleration unit for HDC learning tasks based on the binary spatter-code model. Integrated into the execution stage of the Klessydra-T03 RISC-V core, the unit accelerates the core arithmetic operations of HDC and can be configured at synthesis time in terms of hardware parallelism, supported operations, and size of the local memories, trading off execution time with hardware resources to match application needs. A custom RISC-V Instruction Set Extension efficiently controls the accelerator, with instructions fully integrated into the GNU Compiler Collection toolchain and exposed to the programmer as intrinsics. Dedicated Control and Status Registers allow specifying the characteristics of the high-dimensional space and the target learning tasks at runtime, controlling the hardware loops of the accelerator and enabling the same hardware architecture to be used for various tasks. The dual flexibility coming from hardware configuration and software programmability sets this work apart from application-specific solutions in the literature, offering a unique, versatile accelerator adaptable to a wide range of applications and learning tasks.
- 10 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 75 (2) , 653-664
Configurable Hardware Acceleration for Hyperdimensional Computing Extension on RISC-V
Hyperdimensional Computing (HDC) is a neuro-inspired computational model that represents and manipulates information using high-dimensional distributed representations that are combined and compared using simple and highly parallel vector operations. In this work, we present a highly flexible hardware acceleration unit for HDC learning tasks based on the binary spatter-code model. Integrated into the execution stage of the Klessydra-T03 RISC-V core, the unit accelerates the core arithmetic operations of HDC and can be configured at synthesis time in terms of hardware parallelism, supported operations, and size of the local memories, trading off execution time with hardware resources to match application needs. A custom RISC-V Instruction Set Extension efficiently controls the accelerator, with instructions fully integrated into the GNU Compiler Collection toolchain and exposed to the programmer as intrinsics. Dedicated Control and Status Registers allow specifying the characteristics of the high-dimensional space and the target learning tasks at runtime, controlling the hardware loops of the accelerator and enabling the same hardware architecture to be used for various tasks. The dual flexibility coming from hardware configuration and software programmability sets this work apart from application-specific solutions in the literature, offering a unique, versatile accelerator adaptable to a wide range of applications and learning tasks.
- Open Access
- 1 October 2025
- Published by MDPI AG in Electronics
- Vol. 14 (19) , 3924
Eight-Bit Vector SoftFloat Extension for the RISC-V Spike Simulator
The recent demand for 8-bit floating-point (FP) formats is driven by their potential to accelerate domain-specific applications with intensive vector computations (e.g., machine learning, graphics, and data compression). This paper presents the design, implementation, and application of the software model of an 8-bit FP vector arithmetic operation set, compliant with the RISC-V vector instruction set architecture. The model has been developed as an extension of the SoftFloat library and integrated into the RISC-V reference instruction-level simulator Spike, providing the first open-source 8-bit SoftFloat extension for an instruction-set simulator. Based on the SoftFloat library templates for standard FP formats, the proposed extension implements the two widely used 8-bit formats E4M3 and E5M2 in both Open Compute Project (OCP) and IEEE 754 variants. In host-time micro-kernels, FP8 delivers +2–4% more elements per second versus FP32 (across vfadd/vfsub/vfmul) and ≈5% lower RSS; E4M3 and E5M2 perform similarly. Enabling FP8 in Spike increases the stripped binary by ~1.8% (mostly .text). The proposed extension was used to fully verify and correct errors in the vector FP unit design for the eProcessor European project, and continues to be used to verify other 8-bit FP unit implementations.
- 1 October 2025
- Published by MDPI AG in Electronics
- Vol. 14 (19) , 3924
Open AccessEight-Bit Vector SoftFloat Extension for the RISC-V Spike Simulator
The recent demand for 8-bit floating-point (FP) formats is driven by their potential to accelerate domain-specific applications with intensive vector computations (e.g., machine learning, graphics, and data compression). This paper presents the design, implementation, and application of the software model of an 8-bit FP vector arithmetic operation set, compliant with the RISC-V vector instruction set architecture. The model has been developed as an extension of the SoftFloat library and integrated into the RISC-V reference instruction-level simulator Spike, providing the first open-source 8-bit SoftFloat extension for an instruction-set simulator. Based on the SoftFloat library templates for standard FP formats, the proposed extension implements the two widely used 8-bit formats E4M3 and E5M2 in both Open Compute Project (OCP) and IEEE 754 variants. In host-time micro-kernels, FP8 delivers +2–4% more elements per second versus FP32 (across vfadd/vfsub/vfmul) and ≈5% lower RSS; E4M3 and E5M2 perform similarly. Enabling FP8 in Spike increases the stripped binary by ~1.8% (mostly .text). The proposed extension was used to fully verify and correct errors in the vector FP unit design for the eProcessor European project, and continues to be used to verify other 8-bit FP unit implementations.
- Open Access
- 11 July 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Embedded Computing Systems
- Vol. 24 (4) , 1-23
Efficient Implementation of LinearUCB through Algorithmic Improvements and Vector Computing Acceleration for Embedded Learning Systems
As the Internet of Things expands, embedding Artificial Intelligence algorithms in resource-constrained devices has become increasingly important to enable real-time, autonomous decision-making without relying on centralized cloud servers. However, implementing and executing complex algorithms in embedded devices poses significant challenges due to limited computational power, memory, and energy resources. 2This article presents algorithmic and hardware techniques to efficiently implement two LinearUCB Contextual Bandits algorithms on resource-constrained embedded devices. Algorithmic modifications based on the Sherman–Morrison–Woodbury formula streamline model complexity, while vector acceleration is harnessed to speed up matrix operations. We analyze the impact of each optimization individually and then combine them in a two-pronged strategy. The results show notable improvements in execution time and energy consumption, demonstrating the effectiveness of combining algorithmic and hardware optimizations to enhance learning models for edge computing environments with low-power and real-time requirements.
- 11 July 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Embedded Computing Systems
- Vol. 24 (4) , 1-23
Open AccessEfficient Implementation of LinearUCB through Algorithmic Improvements and Vector Computing Acceleration for Embedded Learning Systems
As the Internet of Things expands, embedding Artificial Intelligence algorithms in resource-constrained devices has become increasingly important to enable real-time, autonomous decision-making without relying on centralized cloud servers. However, implementing and executing complex algorithms in embedded devices poses significant challenges due to limited computational power, memory, and energy resources. 2This article presents algorithmic and hardware techniques to efficiently implement two LinearUCB Contextual Bandits algorithms on resource-constrained embedded devices. Algorithmic modifications based on the Sherman–Morrison–Woodbury formula streamline model complexity, while vector acceleration is harnessed to speed up matrix operations. We analyze the impact of each optimization individually and then combine them in a two-pronged strategy. The results show notable improvements in execution time and energy consumption, demonstrating the effectiveness of combining algorithmic and hardware optimizations to enhance learning models for edge computing environments with low-power and real-time requirements.
- 8 March 2025
- Published by Springer Nature
- p. 56-64
Alternative Implementations of Error Correcting Codes in Register Files of a RISC-V Safety Critical Processor
- 8 March 2025
- Published by Springer Nature
- p. 56-64
Alternative Implementations of Error Correcting Codes in Register Files of a RISC-V Safety Critical Processor
- Open Access
Efficient Implementation of LinearUCB through Algorithmic Improvements and Vector Computing Acceleration for Embedded Learning Systems
As the Internet of Things expands, embedding Artificial Intelligence algorithms in resource-constrained devices has become increasingly important to enable real-time, autonomous decision-making without relying on centralized cloud servers. However, implementing and executing complex algorithms in embedded devices poses significant challenges due to limited computational power, memory, and energy resources. This paper presents algorithmic and hardware techniques to efficiently implement two LinearUCB Contextual Bandits algorithms on resource-constrained embedded devices. Algorithmic modifications based on the Sherman-Morrison-Woodbury formula streamline model complexity, while vector acceleration is harnessed to speed up matrix operations. We analyze the impact of each optimization individually and then combine them in a two-pronged strategy. The results show notable improvements in execution time and energy consumption, demonstrating the effectiveness of combining algorithmic and hardware optimizations to enhance learning models for edge computing environments with low-power and real-time requirements.
Open AccessEfficient Implementation of LinearUCB through Algorithmic Improvements and Vector Computing Acceleration for Embedded Learning Systems
As the Internet of Things expands, embedding Artificial Intelligence algorithms in resource-constrained devices has become increasingly important to enable real-time, autonomous decision-making without relying on centralized cloud servers. However, implementing and executing complex algorithms in embedded devices poses significant challenges due to limited computational power, memory, and energy resources. This paper presents algorithmic and hardware techniques to efficiently implement two LinearUCB Contextual Bandits algorithms on resource-constrained embedded devices. Algorithmic modifications based on the Sherman-Morrison-Woodbury formula streamline model complexity, while vector acceleration is harnessed to speed up matrix operations. We analyze the impact of each optimization individually and then combine them in a two-pronged strategy. The results show notable improvements in execution time and energy consumption, demonstrating the effectiveness of combining algorithmic and hardware optimizations to enhance learning models for edge computing environments with low-power and real-time requirements.
- Open Access
- 9 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 95720-95735
Dynamic Triple Modular Redundancy in Interleaved Hardware Threads: An Alternative Solution to Lockstep Multi-Cores for Fault-Tolerant Systems
Over the years, significant work has been done on high-integrity systems, such as those found in cars, satellites and aircrafts, to minimize the risk that a logic fault causes a system failure, thus having functional safety as a key requirement. In this study, we employ an innovative approach to harness the benefits of both Dual Modular Redundancy and Triple Modular Redundancy techniques within an Interleaved-Multi-Threading microprocessor core, by means of a microarchitecture design capable of dynamically switching from Dual Modular Redundancy to Triple Modular Redundancy in case of faults. We explain the quantitative results obtained from an extensive fault injection simulation campaign on the fault tolerant core compared with its previous version regarding fault tolerant capabilities. The results show that in several application cases the fault resilience improvement and the hardware and timing overhead are better compared to the lockstep-based dual core approach. The proposed technique achieves 98,6% fault mitigation at the expense of only 4 clock cycles for roll-back overhead, with no checkpointing redundancy.
- 9 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 95720-95735
Open AccessDynamic Triple Modular Redundancy in Interleaved Hardware Threads: An Alternative Solution to Lockstep Multi-Cores for Fault-Tolerant Systems
Over the years, significant work has been done on high-integrity systems, such as those found in cars, satellites and aircrafts, to minimize the risk that a logic fault causes a system failure, thus having functional safety as a key requirement. In this study, we employ an innovative approach to harness the benefits of both Dual Modular Redundancy and Triple Modular Redundancy techniques within an Interleaved-Multi-Threading microprocessor core, by means of a microarchitecture design capable of dynamically switching from Dual Modular Redundancy to Triple Modular Redundancy in case of faults. We explain the quantitative results obtained from an extensive fault injection simulation campaign on the fault tolerant core compared with its previous version regarding fault tolerant capabilities. The results show that in several application cases the fault resilience improvement and the hardware and timing overhead are better compared to the lockstep-based dual core approach. The proposed technique achieves 98,6% fault mitigation at the expense of only 4 clock cycles for roll-back overhead, with no checkpointing redundancy.
- Open Access
- 8 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 73 (7) , 1767-1779
Design, Implementation and Evaluation of a New Variable Latency Integer Division Scheme
Integer division is key for various applications and often represents the performance bottleneck due to its inherent mathematical properties that limit its parallelization. This paper presents a new data-dependent variable latency division algorithm derived from the classic non-performing restoring method. The proposed technique exploits the relationship between the number of leading zeros in the divisor and in the partial remainder to dynamically detect and skip those iterations that result in a simple left shift. While a similar principle has been exploited in previous works, the proposed approach outperforms existing variable latency divider schemes in average latency and power consumption. We detail the algorithm and its implementation in four variants, offering versatility for the specific application requirements. For each variant, we report the average latency evaluated with different benchmarks, and we analyze the synthesis results for both FPGA and ASIC deployment, reporting clock speed, average execution time, hardware resources, and energy consumption, compared with existing fixed and variable latency dividers.
- 8 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 73 (7) , 1767-1779
Open AccessDesign, Implementation and Evaluation of a New Variable Latency Integer Division Scheme
Integer division is key for various applications and often represents the performance bottleneck due to its inherent mathematical properties that limit its parallelization. This paper presents a new data-dependent variable latency division algorithm derived from the classic non-performing restoring method. The proposed technique exploits the relationship between the number of leading zeros in the divisor and in the partial remainder to dynamically detect and skip those iterations that result in a simple left shift. While a similar principle has been exploited in previous works, the proposed approach outperforms existing variable latency divider schemes in average latency and power consumption. We detail the algorithm and its implementation in four variants, offering versatility for the specific application requirements. For each variant, we report the average latency evaluated with different benchmarks, and we analyze the synthesis results for both FPGA and ASIC deployment, reporting clock speed, average execution time, hardware resources, and energy consumption, compared with existing fixed and variable latency dividers.
- Open Access
- 16 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 30495-30506
A RISC-V Fault-Tolerant Soft-Processor Based on Full/Partial Heterogeneous Dual-Core Protection
The low probability of single event upsets (SEU) within particular satellite orbits, makes Commercial-off-the-shelf (COTS) electronic components a viable solution for space system implementation, thanks to the introduction of design-level fault tolerance techniques at the expense of some performance/energy/area penalty. This paper illustrates the design and validation of a novel RISC-V dual-core architecture, based on a computing paradigm that we refer to as full/partial heterogeneous multi-core protection. The approach relies on a small, low-performance, fully fault-tolerant core (LP core) coupled with a high-performance partially fault-tolerant core (HP core). The computing paradigm assumes the failure-exposed HP core executes computation intensive routines for relatively short periods of time, making the occurrence of failures a statistically unlikely situation, while the fully fault-tolerant LP core operates in critical control tasks and manages the failure recovery of the high-performance core. The execution time percentage in the LP core varies from a minimum of 11.4% up to a maximum of 91.3% while in the HP core it is between 8.7% and 88.6%, depending on the application. In the proposed study, both the cores belong to the RISC-V compliant Klessydra core family. The dual-core architecture also includes a watchdog timer controlled by the LP core and monitoring the non-protected HP core, and a context switch FIFO that speeds up the code and data switch between the two cores during failure recovery. A dedicated run-time software environment coordinates the execution of tasks on the high-performance core in a resilient fashion. The dual-core processor has been validated through extensive RTL simulations running in an UVM-based fault-injection environment, which emulates SEUs at various rates. Experimental results illustrate the benefits and limits obtained by using a heterogeneous architecture with different levels of protection and performance. The failure probability assuming a SEU fault occurrence can be reduced by a factor between 10X and 30X with respect to the non-protected architecture, leading to an average failure rate of up to 4.00E-06 per second with respect to 1.80E-05 per second in the non-protected architecture.
- 16 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 30495-30506
Open AccessA RISC-V Fault-Tolerant Soft-Processor Based on Full/Partial Heterogeneous Dual-Core Protection
The low probability of single event upsets (SEU) within particular satellite orbits, makes Commercial-off-the-shelf (COTS) electronic components a viable solution for space system implementation, thanks to the introduction of design-level fault tolerance techniques at the expense of some performance/energy/area penalty. This paper illustrates the design and validation of a novel RISC-V dual-core architecture, based on a computing paradigm that we refer to as full/partial heterogeneous multi-core protection. The approach relies on a small, low-performance, fully fault-tolerant core (LP core) coupled with a high-performance partially fault-tolerant core (HP core). The computing paradigm assumes the failure-exposed HP core executes computation intensive routines for relatively short periods of time, making the occurrence of failures a statistically unlikely situation, while the fully fault-tolerant LP core operates in critical control tasks and manages the failure recovery of the high-performance core. The execution time percentage in the LP core varies from a minimum of 11.4% up to a maximum of 91.3% while in the HP core it is between 8.7% and 88.6%, depending on the application. In the proposed study, both the cores belong to the RISC-V compliant Klessydra core family. The dual-core architecture also includes a watchdog timer controlled by the LP core and monitoring the non-protected HP core, and a context switch FIFO that speeds up the code and data switch between the two cores during failure recovery. A dedicated run-time software environment coordinates the execution of tasks on the high-performance core in a resilient fashion. The dual-core processor has been validated through extensive RTL simulations running in an UVM-based fault-injection environment, which emulates SEUs at various rates. Experimental results illustrate the benefits and limits obtained by using a heterogeneous architecture with different levels of protection and performance. The failure probability assuming a SEU fault occurrence can be reduced by a factor between 10X and 30X with respect to the non-protected architecture, leading to an average failure rate of up to 4.00E-06 per second with respect to 1.80E-05 per second in the non-protected architecture.
- 13 January 2024
- Published by Springer Nature
- p. 36-41
A Universal Hardware Emulator for Verification IPs on FPGA: A Novel and Low-Cost Approach
- 13 January 2024
- Published by Springer Nature
- p. 36-41
A Universal Hardware Emulator for Verification IPs on FPGA: A Novel and Low-Cost Approach
- 13 January 2024
- Published by Springer Nature
- p. 42-48
Single Event Transient Reliability Analysis on a Fault-Tolerant RISC-V Microprocessor Design
- 13 January 2024
- Published by Springer Nature
- p. 42-48
Single Event Transient Reliability Analysis on a Fault-Tolerant RISC-V Microprocessor Design