Browsing Τμήμα Ηλεκτρολόγων Μηχανικών και Μηχανικών Υπολογιστών / Department of Electrical and Computer Engineering by Author "265204fe-37ad-4475-8508-c8b586ba7f45"
Now showing items 1-8 of 8
-
Article
ArrayFlex: A Systolic Array Architecture with Configurable Transparent Pipelining
Peltekis, Christodoulos; Filippas, Dionysios; Dimitrakopoulos, Giorgos; Nicopoulos, Chrysostomos; Pnevmatikatos, Dionisios (IEEE, 2023-06-02)Convolutional Neural Networks (CNNs) are the state-of-the-art solution for many deep learning applications. For maximum scalability, their computation should combine high performance and energy efficiency. In practice, the ...
-
Article
The Case for Asymmetric Systolic Array Floorplanning
Peltekis, Christodoulos; Filippas, Dionysios; Dimitrakopoulos, Giorgos; Nicopoulos, Chrysostomos (IEEE, 2023-09)The widespread proliferation of deep learning applications has triggered the need to accelerate them directly in hardware. General Matrix Multiplication (GEMM) kernels are elemental deep-learning constructs and they ...
-
Article
Exploiting data encoding and reordering for low-power streaming in systolic arrays
Peltekis, Christodoulos; Filippas, Dionysios; Dimitrakopoulos, Giorgos; Nicopoulos, Chrysostomos (Elsevier, 2023-10-03)Systolic Array (SA) architectures are well-suited for accelerating matrix multiplications through the use of a pipelined array of Processing Elements (PEs) communicating with local connections and pre-orchestrated data ...
-
Article
IndexMAC: A Custom RISC-V Vector Instruction to Accelerate Structured-Sparse Matrix Multiplications
Titopoulos, Vasileios; Alexandridis, Kosmas; Peltekis, Christodoulos; Nicopoulos, Chrysostomos; Dimitrakopoulos, Giorgos (IEEE, 2024-03)Structured sparsity has been proposed as an efficient way to prune the complexity of modern Machine Learning (ML) applications and to simplify the handling of sparse data in hardware. The acceleration of ML models - for ...
-
Article
Low-Power Data Streaming in Systolic Arrays with Bus-Invert Coding and Zero-Value Clock Gating
Peltekis, Christodoulos; Filippas, Dionysios; Dimitrakopoulos, Giorgos; Nicopoulos, Chrysostomos (IEEE, 2023-07-17)Systolic Array (SA) architectures are well suited for accelerating matrix multiplications through the use of a pipelined array of Processing Elements (PEs) communicating with local connections and pre-orchestrated data ...
-
Article
Multi-Armed Bandits for Autonomous Test Application in RISC-V Processor Verification
Dimitrakopoulos, Giorgos; Kallitsounakis, E.; Takakis, Zacharias; Stefanidis, Apostolos; Nicopoulos, Chrysostomos (IEEE, 2023-07-17)Multi-armed bandit problems have recently received a great deal of attention, because they adequately formalize so called exploration-exploitation trade-offs arising in several relevant applications of recommendation ...
-
Article
Reduced-Precision Floating-Point Arithmetic in Systolic Arrays with Skewed Pipelines
Filippas, Dionysios; Peltekis, Christodoulos; Dimitrakopoulos, Giorgos; Nicopoulos, Chrysostomos (IEEE, 2023-07-07)The acceleration of deep-learning kernels in hardware relies on matrix multiplications that are executed efficiently on Systolic Arrays (SA). To effectively trade off deep-learning training/inference quality with hardware ...
-
Article
Streaming Dilated Convolution Engine
Filippas, Dionysios; Nicopoulos, Chrysostomos; Dimitrakopoulos, Giorgos (IEEE, 2023-01-09)Convolution is one of the most critical operations in various application domains and its computation should combine high performance with energy efficiency. This requirement is critical both for standard convolution and ...