FPGA-Based Accelerators for Deep Neural Networks

Dear Colleagues,

In recent years, deep learning has achieved remarkable breakthroughs across various artificial intelligence (AI) domains, including computer vision, natural language processing, and generative AI. The advent of large-scale models, such as Transformer-based Large Language Models (LLMs) and Diffusion models, has further pushed the boundaries of AI capabilities. However, these models come with ever-increasing computational complexity and memory demands, posing significant challenges to conventional computing platforms in terms of performance, energy efficiency, and scalability. Neuromorphic computing, inspired by the brain's neural architecture, offers a promising pathway toward ultra-low-power intelligent systems. In this context, Field-Programmable Gate Arrays (FPGAs) have emerged as a highly attractive platform for accelerating both deep learning and neuromorphic algorithms, from edge devices to cloud servers. Key advantages of FPGAs include their high reconfigurability, rapid deployment cycles, capability for customized architecture design, and support for software–hardware co-design in System-on-Chip (SoC) configurations.

This Special Issue aims to showcase cutting-edge research on hardware acceleration of deep neural networks using FPGAs, with particular interest in optimizations for modern architectures like Transformers, LLMs, and Diffusion models. Topics of interest include, but are not limited to, the following:

• Algorithm–hardware co-design for efficient FPGA-based DNN acceleration;
• System-level design and software tools for compiling and deploying models on FPGAs;
• Reconfigurable and adaptive computing architectures for AI/ML workloads;
• FPGA-based rapid prototyping of ML systems, including large-scale models;
• Programmable neuromorphic and spiking neural network implementations on FPGAs;
• Deployment and evaluation of Transformer, LLM, and Diffusion models on reconfigurable hardware;
• Design of FPGA accelerators with optimized attention mechanisms and generative model blocks;
• AI systems based on coarse-grained reconfigurable architectures (CGRAs);
• Novel applications and case studies demonstrating FPGA-based intelligence.

We invite original contributions that address the optimization, implementation, and application of FPGA-based accelerators for current and next-generation deep learning models.

Dr. Yufei Ma
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Electronics is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.