Konstantinos Iliakis
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
Konstantinos Iliakis has a total of 44 co-authors. Here are the top co-authors by number of shared publications:
- Dimitrios Soudris
National Technical University of Athens
6 shared publications - Sotirios Xydis
National Technical University of Athens
5 shared publications Helga Timko
2 shared publicationsPanagiotis Tsapatsaris
2 shared publicationsAlexandre Lasheen
1 shared publication- Alexandre Lasheen
CERN
1 shared publication
Latest Publications
- Open Access
- 27 June 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 23 (2) , 1-25
AccelHSA: Modeling Single-ISA Heterogeneous GPU Architectures
Initially branded as dedicated graphics processing accelerators, GPUs now find applications in an ever-growing range of domains, including artificial intelligence, high-performance computing, self-driving vehicles, and bioinformatics. However, this diversity comes at the cost of reduced resource efficiency and micro-architectural affinity. Evidently, the homogeneity of the GPU hardware struggles to cope with the vast heterogeneity of GPU applications. Motivated by the aforementioned observations, this article introduces the concept of Single-ISA Heterogeneous GPU architectures. In order to explore the efficiency of the new GPU architectural paradigm, we extend Accel-Sim, the state-of-the-art, cycle-accurate GPU simulator to support single-ISA heterogeneous cores within the GPU chip. The proposed implementation, called AccelHSA, supports independently tuning the micro-architectural characteristics of the cores, unlocking a wide design space. The CUDA API is extended to allow control of the kernel-to-core-type mapping along with a newly developed kernel launching model that supports concurrent execution, aimed at, albeit not limited to, the context of the simulator. We showcase the impact of single-ISA heterogeneous GPU architectures via a case study targeting the collocation of resource sensitive and insensitive HPC kernels. Finally, the heterogeneous GPU architecture is evaluated against homogeneous GPU baselines, demonstrating a 27.07% average speedup with a marginal 0.47% area overhead.
- 27 June 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 23 (2) , 1-25
Open AccessAccelHSA: Modeling Single-ISA Heterogeneous GPU Architectures
Initially branded as dedicated graphics processing accelerators, GPUs now find applications in an ever-growing range of domains, including artificial intelligence, high-performance computing, self-driving vehicles, and bioinformatics. However, this diversity comes at the cost of reduced resource efficiency and micro-architectural affinity. Evidently, the homogeneity of the GPU hardware struggles to cope with the vast heterogeneity of GPU applications. Motivated by the aforementioned observations, this article introduces the concept of Single-ISA Heterogeneous GPU architectures. In order to explore the efficiency of the new GPU architectural paradigm, we extend Accel-Sim, the state-of-the-art, cycle-accurate GPU simulator to support single-ISA heterogeneous cores within the GPU chip. The proposed implementation, called AccelHSA, supports independently tuning the micro-architectural characteristics of the cores, unlocking a wide design space. The CUDA API is extended to allow control of the kernel-to-core-type mapping along with a newly developed kernel launching model that supports concurrent execution, aimed at, albeit not limited to, the context of the simulator. We showcase the impact of single-ISA heterogeneous GPU architectures via a case study targeting the collocation of resource sensitive and insensitive HPC kernels. Finally, the heterogeneous GPU architecture is evaluated against homogeneous GPU baselines, demonstrating a 27.07% average speedup with a marginal 0.47% area overhead.
- Open Access
- 28 May 2025
- Published by Association for Computing Machinery (ACM)
- p. 228-229
POSTER: Performance Portability in GPU-Accelerated Spectral Finite Element Fluid Simulations: A Cross-layer Exploration Approach
- 28 May 2025
- Published by Association for Computing Machinery (ACM)
- p. 228-229
Open AccessPOSTER: Performance Portability in GPU-Accelerated Spectral Finite Element Fluid Simulations: A Cross-layer Exploration Approach
- Open Access
- 29 November 2023
- Published by American Physical Society (APS) in Physical Review Accelerators and Beams
- Vol. 26 (11) , 114602
Beam longitudinal dynamics simulation studies
The beam longitudinal dynamics code blond, utilized tool, has been developed at CERN since 2014. It has emerged as a central tool for conducting longitudinal beam dynamics simulations. In this paper, we present this modular simulation suite and the various physics models that can be included and combined by the user. We detail the reference frame, the equations of motion, the plethora of options for radio-frequency parameters such as phase noise, fixed-field acceleration, and feedback models for the CERN accelerators, as well as the modeling of collective effects and synchrotron radiation. We also present various methods of generating multibunch distributions matched to a given impedance model. blond is furthermore a well tested and optimized simulation suite, which is demonstrated through examples, too.
- 29 November 2023
- Published by American Physical Society (APS) in Physical Review Accelerators and Beams
- Vol. 26 (11) , 114602
Open AccessBeam longitudinal dynamics simulation studies
The beam longitudinal dynamics code blond, utilized tool, has been developed at CERN since 2014. It has emerged as a central tool for conducting longitudinal beam dynamics simulations. In this paper, we present this modular simulation suite and the various physics models that can be included and combined by the user. We detail the reference frame, the equations of motion, the plethora of options for radio-frequency parameters such as phase noise, fixed-field acceleration, and feedback models for the CERN accelerators, as well as the modeling of collective effects and synchrotron radiation. We also present various methods of generating multibunch distributions matched to a given impedance model. blond is furthermore a well tested and optimized simulation suite, which is demonstrated through examples, too.
- Open Access
Beam Longitudinal Dynamics Simulation Suite BLonD
The beam longitudinal dynamics code BLonD has been developed at CERN since 2014 and has become a central tool for longitudinal beam dynamics simulations. In this paper, we present this modular simulation suite and the various physics models that can be included and combined by the user. We detail the reference frame, the equations of motion, the BLonD-specific options for radio-frequency parameters such as phase noise, fixed-field acceleration, and feedback models for the CERN accelerators, as well as the modeling of collective effects and synchrotron radiation. We also present various methods of generating multi-bunch distributions matched to a given impedance model. BLonD is furthermore a well-tested and optimized simulation suite, which is demonstrated through examples, too.
Open AccessBeam Longitudinal Dynamics Simulation Suite BLonD
The beam longitudinal dynamics code BLonD has been developed at CERN since 2014 and has become a central tool for longitudinal beam dynamics simulations. In this paper, we present this modular simulation suite and the various physics models that can be included and combined by the user. We detail the reference frame, the equations of motion, the BLonD-specific options for radio-frequency parameters such as phase noise, fixed-field acceleration, and feedback models for the CERN accelerators, as well as the modeling of collective effects and synchrotron radiation. We also present various methods of generating multi-bunch distributions matched to a given impedance model. BLonD is furthermore a well-tested and optimized simulation suite, which is demonstrated through examples, too.
- Open Access
- 20 July 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 33 (12) , 4425-4439
Enabling Large Scale Simulations for Particle Accelerators
International high-energy particle physics research centers, like CERN and Fermilab, require excessive studies and simulations to plan for the upcoming upgrades of the world's largest particle accelerators, and the design of future machines given the technological challenges and tight budgetary constraints. The Beam Longitudinal Dynamics (BLonD) simulator suite incorporates the most detailed and complex physics phenomena in the field of longitudinal beam dynamics, required for providing extremely accurate predictions. Modern challenges in beam dynamics dictate for longer, larger and numerous simulation studies to draw meaningful conclusions that will drive the baseline choices for the daily operation of current machines and the design choices of future projects. These studies are extremely time consuming, and would be impractical to perform without a High-Performance Computing oriented simulator framework. In this article, at first, we design and evaluate a highly-optimized distributed version of BLonD. We combine approximate computing techniques, and leverage a dynamic load-balancing scheme to relax synchronization and improve scalability. In addition, we employ GPUs to accelerate the distributed implementation. We evaluate the highly optimized distributed beam longitudinal dynamics simulator in a supercomputing system and demonstrate speedups of more than two orders of magnitude when run on 32 GPU platforms, w.r.t. the previous state-of-art. By driving a wide range of new studies, the proposed high performance beam longitudinal dynamics simulator forms an invaluable tool for accelerator physicists.
- 20 July 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 33 (12) , 4425-4439
Open AccessEnabling Large Scale Simulations for Particle Accelerators
International high-energy particle physics research centers, like CERN and Fermilab, require excessive studies and simulations to plan for the upcoming upgrades of the world's largest particle accelerators, and the design of future machines given the technological challenges and tight budgetary constraints. The Beam Longitudinal Dynamics (BLonD) simulator suite incorporates the most detailed and complex physics phenomena in the field of longitudinal beam dynamics, required for providing extremely accurate predictions. Modern challenges in beam dynamics dictate for longer, larger and numerous simulation studies to draw meaningful conclusions that will drive the baseline choices for the daily operation of current machines and the design choices of future projects. These studies are extremely time consuming, and would be impractical to perform without a High-Performance Computing oriented simulator framework. In this article, at first, we design and evaluate a highly-optimized distributed version of BLonD. We combine approximate computing techniques, and leverage a dynamic load-balancing scheme to relax synchronization and improve scalability. In addition, we employ GPUs to accelerate the distributed implementation. We evaluate the highly optimized distributed beam longitudinal dynamics simulator in a supercomputing system and demonstrate speedups of more than two orders of magnitude when run on 32 GPU platforms, w.r.t. the previous state-of-art. By driving a wide range of new studies, the proposed high performance beam longitudinal dynamics simulator forms an invaluable tool for accelerator physicists.
- Open Access
- 4 March 2022
- Published by Institution of Engineering and Technology (IET) in IET Blockchain
- Vol. 2 (1) , 1-12
GPU accelerated blockchain over key‐value database transactions
Blockchain is a distributed ledger based on peer‐to‐peer networks, originally used for crypto‐currency systems. Blockchains are being used as an enabling technology for decentralised applications in the areas of Internet‐of‐Things, finance, supply‐chain and others. Consistency, data privacy, performance, and energy efficiency are of paramount importance in such applications. The full nodes of public, permissionless blockchains undertake the task of verifying the transactions generated by the network. Full nodes perform operations such as confirming balances, transactions, and history, i.e. mostly database search queries. Consequently, their throughput is crucial for the performance of blockchain systems. In this work, the benefits of accelerating the blockchain search and insert queries by leveraging GPU platforms are studied. An extensive comparison between the most dominant utilized database is provided, i.e. LevelDB, and MegaKV, a high‐performance GPU‐accelerated database. Realistic operations that take place in blockchain systems are emulated and evaluated over representative scenarios, showing three orders of magnitude gains in throughput and energy efficiency without compromising the security aspect. The extensive comparison between LevelDB and MegaKV indicates that GPU acceleration is an effective solution for runtime and energy efficiency enhancement of blockchain systems, and the integration of the two technologies is a promising field of research.
- 4 March 2022
- Published by Institution of Engineering and Technology (IET) in IET Blockchain
- Vol. 2 (1) , 1-12
Open AccessGPU accelerated blockchain over key‐value database transactions
Blockchain is a distributed ledger based on peer‐to‐peer networks, originally used for crypto‐currency systems. Blockchains are being used as an enabling technology for decentralised applications in the areas of Internet‐of‐Things, finance, supply‐chain and others. Consistency, data privacy, performance, and energy efficiency are of paramount importance in such applications. The full nodes of public, permissionless blockchains undertake the task of verifying the transactions generated by the network. Full nodes perform operations such as confirming balances, transactions, and history, i.e. mostly database search queries. Consequently, their throughput is crucial for the performance of blockchain systems. In this work, the benefits of accelerating the blockchain search and insert queries by leveraging GPU platforms are studied. An extensive comparison between the most dominant utilized database is provided, i.e. LevelDB, and MegaKV, a high‐performance GPU‐accelerated database. Realistic operations that take place in blockchain systems are emulated and evaluated over representative scenarios, showing three orders of magnitude gains in throughput and energy efficiency without compromising the security aspect. The extensive comparison between LevelDB and MegaKV indicates that GPU acceleration is an effective solution for runtime and energy efficiency enhancement of blockchain systems, and the integration of the two technologies is a promising field of research.
- 29 June 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 33 (2) , 388-402
Repurposing GPU Microarchitectures with Light-Weight Out-Of-Order Execution
GPU is the dominant platform for accelerating general-purpose workloads due to its computing capacity and cost-efficiency. GPU applications cover an ever-growing range of domains. To achieve high throughput, GPUs rely on massive multi-threading and fast context switching to overlap computations with memory operations. We observe that among the diverse GPU workloads, there exists a significant class of kernels that fail to maintain a sufficient number of active warps to hide the latency of memory operations, and thus suffer from frequent stalling. We argue that the dominant Thread-Level Parallelism model is not enough to efficiently accommodate the variability of modern GPU applications. To address this inherent inefficiency, we propose a novel micro-architecture with lightweight Out-Of-Order execution capability enabling Instruction-Level Parallelism to complement the conventional Thread-Level Parallelism model. To minimize the hardware overhead, we carefully design our extension to highly re-use the existing micro-architectural structures and study various design trade-offs to contain the overall area and power overhead, while providing improved performance. We show that the proposed architecture outperforms traditional platforms by 23% on average for low-occupancy kernels, with an area and power overhead of 1.29% and 10.05%, respectively. Finally, we establish the potential of our proposal as a micro-architecture alternative by providing 16\% speedup over a wide collection of 60 general-purpose kernels.
- 29 June 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 33 (2) , 388-402
Repurposing GPU Microarchitectures with Light-Weight Out-Of-Order Execution
GPU is the dominant platform for accelerating general-purpose workloads due to its computing capacity and cost-efficiency. GPU applications cover an ever-growing range of domains. To achieve high throughput, GPUs rely on massive multi-threading and fast context switching to overlap computations with memory operations. We observe that among the diverse GPU workloads, there exists a significant class of kernels that fail to maintain a sufficient number of active warps to hide the latency of memory operations, and thus suffer from frequent stalling. We argue that the dominant Thread-Level Parallelism model is not enough to efficiently accommodate the variability of modern GPU applications. To address this inherent inefficiency, we propose a novel micro-architecture with lightweight Out-Of-Order execution capability enabling Instruction-Level Parallelism to complement the conventional Thread-Level Parallelism model. To minimize the hardware overhead, we carefully design our extension to highly re-use the existing micro-architectural structures and study various design trade-offs to contain the overall area and power overhead, while providing improved performance. We show that the proposed architecture outperforms traditional platforms by 23% on average for low-occupancy kernels, with an area and power overhead of 1.29% and 10.05%, respectively. Finally, we establish the potential of our proposal as a micro-architecture alternative by providing 16\% speedup over a wide collection of 60 general-purpose kernels.
- 4 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Architecture Letters
- Vol. 18 (2) , 166-169
LOOG: Improving GPU Efficiency With Light-Weight Out-Of-Order Execution
GPUs are one of the most prevalent platforms for accelerating general-purpose workloads due to their intuitive programming model, computing capacity, and cost-effectiveness. GPUs rely on massive multi-threading and fast context switching to overlap computations with memory operations. Among the diverse GPU workloads, there exists a class of kernels that fail to maintain a sufficient number of active warps to hide the latency of memory operations, and thus suffer from frequent stalling. We observe that these kernels will benefit from increased levels of Instruction-Level Parallelism and we propose a novel architecture with lightweight Out-Of-Order execution capability. To minimize hardware overheads, we carefully design our extension to highly re-use the existing micro-architectural structures. We show that the proposed architecture outperforms traditional platforms by 15% to 46% on average for low occupancy kernels, with an area overhead of 0.74% to 3.94%. Finally, we prove the potential of our proposal as a GPU u-arch alternative, by providing a 5% speedup over a wide collection of 63 general-purpose kernels with as little as 0.74% area overhead.
- 4 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Architecture Letters
- Vol. 18 (2) , 166-169
LOOG: Improving GPU Efficiency With Light-Weight Out-Of-Order Execution
GPUs are one of the most prevalent platforms for accelerating general-purpose workloads due to their intuitive programming model, computing capacity, and cost-effectiveness. GPUs rely on massive multi-threading and fast context switching to overlap computations with memory operations. Among the diverse GPU workloads, there exists a class of kernels that fail to maintain a sufficient number of active warps to hide the latency of memory operations, and thus suffer from frequent stalling. We observe that these kernels will benefit from increased levels of Instruction-Level Parallelism and we propose a novel architecture with lightweight Out-Of-Order execution capability. To minimize hardware overheads, we carefully design our extension to highly re-use the existing micro-architectural structures. We show that the proposed architecture outperforms traditional platforms by 15% to 46% on average for low occupancy kernels, with an area overhead of 0.74% to 3.94%. Finally, we prove the potential of our proposal as a GPU u-arch alternative, by providing a 5% speedup over a wide collection of 63 general-purpose kernels with as little as 0.74% area overhead.
- 17 April 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Architecture Letters
- Vol. 17 (2) , 143-146
Decoupled MapReduce for Shared-Memory Multi-Core Architectures
Modern multi-core processors exhibit high integration densities, e.g. up to several tens of cores. Multiple programming frameworks have emerged to facilitate the development of highly parallel applications. The MapReduce programming model, after having demonstrated its usability in the area of distributed computing systems, has been adapted to the needs of shared-memory multi-processors showing also promising results in comparison with conventional multi-threaded libraries, e.g. pthreads. In this paper we enhance the traditional MapReduce architecture by decoupling the map and combine phases in order to boost parallel execution. We show that combiners' memory intensive features limit the system's degree of parallelism, thus resulting in sub-optimal hardware utilization, leaving space for further performance improvements. The proposed decoupled MapReduce architecture is evaluated into a NUMA server platform, showing that the adoption of the proposed De-MapR runtime enables more efficient hardware utilization and competent run-time improvements. We demonstrate that the proposed solution achieves execution speedups of up to 2.46x compared to a state-of-the-art, shared-memory MapReduce library.
- 17 April 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Architecture Letters
- Vol. 17 (2) , 143-146
Decoupled MapReduce for Shared-Memory Multi-Core Architectures
Modern multi-core processors exhibit high integration densities, e.g. up to several tens of cores. Multiple programming frameworks have emerged to facilitate the development of highly parallel applications. The MapReduce programming model, after having demonstrated its usability in the area of distributed computing systems, has been adapted to the needs of shared-memory multi-processors showing also promising results in comparison with conventional multi-threaded libraries, e.g. pthreads. In this paper we enhance the traditional MapReduce architecture by decoupling the map and combine phases in order to boost parallel execution. We show that combiners' memory intensive features limit the system's degree of parallelism, thus resulting in sub-optimal hardware utilization, leaving space for further performance improvements. The proposed decoupled MapReduce architecture is evaluated into a NUMA server platform, showing that the adoption of the proposed De-MapR runtime enables more efficient hardware utilization and competent run-time improvements. We demonstrate that the proposed solution achieves execution speedups of up to 2.46x compared to a state-of-the-art, shared-memory MapReduce library.