Architectural Design and Performance Analysis of FPGA based AI Accelerators: A Comprehensive Review
Not on the map. Why?
Abstract (arXiv)
Deep learning (DL) has emerged as a rapidly developing advanced technology, enabling the performance of complex tasks involving image recognition, natural language processing, and autonomous decision-making with high levels of accuracy. However, as these technologies evolve and strive to meet the growing demands of real-life applications, the complexity of DL models continues to increase. These models require processing of massive volumes of data, demanding substantial computational power and memory bandwidth. This gives rise to the critical need for hardware accelerators that can deliver both high performance and energy efficiency. Accelerator types include ASIC based solutions, GPU accelerators, and FPGA based implementations. The limitations of ASIC and GPU accelerators have led to FPGAs becoming one of the prominent solutions, offering distinct advantages for DL workloads. FPGAs provide a flexible and reconfigurable platform, allowing model specific customization while maintaining high efficiency. This article explores various hardware level optimizations for DL. These optimizations include techniques such as loop pipelining, parallelism, quantization, and various memory hierarchy enhancements. In addition, it provides an overview of state-of-the-art FPGA-based neural network accelerators. Through the study and analysis of these accelerators, several challenges have been identified, paving the way for future optimizations and innovations in the design of FPGA-based hardware accelerators.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 33 of its 148 references that are in the map. First posted on arXiv 25 Feb 2026.
Builds on
33 of 148 references.
References · 33
- 2017In-Datacenter Performance Analysis of a Tensor Processing Unitcited by 618
- 2014Speeding-up Convolutional Neural Networks Using Fine-tuned CP-Decompositioncited by 203
- 2019Bit-Flip Attack: Crushing Neural Network with Progressive Bit Searchcited by 108
- 2018NVIDIA Tensor Core Programmability, Performance & Precisioncited by 79
- 2016ESE: Efficient Speech Recognition Engine with Sparse LSTM on FPGAcited by 52
- 2019FPGA-based Accelerators of Deep Learning Networks for Learning and Classification: A Reviewcited by 29
- 2018Toolflows for Mapping Convolutional Neural Networks on FPGAs: A Survey and Future Directionscited by 27
- 2020EdgeDRNN: Recurrent Neural Network Accelerator for Edge Inferencecited by 19
- 2020A Deep Attentive Convolutional Neural Network for Automatic Cortical Plate Segmentation in Fetal MRIcited by 14
- 2017An OpenCL(TM) Deep Learning Accelerator on Arria 10cited by 13
- 2021GCoD: Graph Convolutional Network Acceleration via Dedicated Algorithm and Accelerator Co-Designcited by 13
- 2023FireFly: A High-Throughput Hardware Accelerator for Spiking Neural Networks with Efficient DSP and Memory Optimizationcited by 13
21 more
- 2019Randomized Tensor Ring Decomposition and Its Application to Large-scale Data Reconstructioncited by 11
- 2021BlockGNN: Towards Efficient GNN Acceleration Using Block-Circulant Weight Matricescited by 11
- 2023Hier-RTLMP: A Hierarchical Automatic Macro Placer for Large-scale Complex IP Blockscited by 11
- 2023Using the IBM Analog In-Memory Hardware Acceleration Kit for Neural Network Training and Inferencecited by 11
- 2018FINN-L: Library Extensions and Design Trade-off Analysis for Variable Precision LSTM Networks on FPGAscited by 10
- 2019Automatic Compiler Based FPGA Accelerator for CNN Trainingcited by 7
- 2022Skydiver: A Spiking Neural Network Accelerator Exploiting Spatio-Temporal Workload Balancecited by 5
- 2023DeepFire2: A Convolutional Spiking Neural Network Accelerator on FPGAscited by 4
- 2024DG-RePlAce: A Dataflow-Driven GPU-Accelerated Analytical Global Placement Framework for Machine Learning Acceleratorscited by 4
- 2024Deep Convolutional Neural Networks Meet Variational Shape Compactness Priors for Image Segmentationcited by 4
- 2024FireFly-S: Exploiting Dual-Side Sparsity for Spiking Neural Networks Acceleration with Reconfigurable Spatial Architecturecited by 3
- 2023Dynasparse: Accelerating GNN Inference through Dynamic Sparsity Exploitationcited by 2
- 2025DASH: A Meta-Attack Framework for Synthesizing Effective and Stealthy Adversarial Examplescited by 2
- 2017CNN-MERP: An FPGA-Based Memory-Efficient Reconfigurable Processor for Forward and Backward Propagation of Convolutional Neural Networkscited by 1
- 2019AccUDNN: A GPU Memory Efficient Accelerator for Training Ultra-deep Neural Networkscited by 1
- 2024Hardware-Software Co-optimised Fast and Accurate Deep Reconfigurable Spiking Inference Accelerator Architecture Design Methodologycited by 1
- 2025Runtime Detection of Adversarial Attacks in AI Accelerators Using Performance Counterscited by 1
- 2025Hardware Efficient Accelerator for Spiking Transformer With Reconfigurable Parallel Time Step Computingcited by 1
- 2025Secure and Storage-Efficient Deep Learning Models for Edge AI Using Automatic Weight Generationcited by 1
- 2025LEGO: Spatial Accelerator Generation and Optimization for Tensor Applicationscited by 1
- 2025A Scalable FPGA Architecture With Adaptive Memory Utilization for GEMM-Based Operationscited by 1
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2025The Role of Advanced Computer Architectures in Accelerating Artificial Intelligence Workloads
- 2022Hardware-Efficient Template-Based Deep CNNs Accelerator Design ↗
- 2024A survey on FPGA-based accelerator for ML models
- 2025Real Time FPGA Based CNNs for Detection, Classification, and Tracking in Autonomous Systems: State of the Art Designs and Optimizations ↗
- 2026A Reconfigurable Framework for AI-FPGA Agent Integration and Acceleration ↗
- 2022Design of High-Throughput Mixed-Precision CNN Accelerators on FPGA
- 2021Reconfigurable co-processor architecture with limited numerical precision to accelerate deep convolutional neural networks
- 2025Energy-Efficient FPGA Framework for Non-Quantized Convolutional Neural Networks