A Parameterizable Convolution Accelerator for Embedded Deep Learning Applications
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918322002460672 |
|---|---|
| author | Mousouliotis, Panagiotis Keramidas, Georgios |
| author_facet | Mousouliotis, Panagiotis Keramidas, Georgios |
| contents | Convolutional neural network (CNN) accelerators implemented on Field-Programmable Gate Arrays (FPGAs) are typically designed with a primary focus on maximizing performance, often measured in giga-operations per second (GOPS). However, real-life embedded deep learning (DL) applications impose multiple constraints related to latency, power consumption, area, and cost. This work presents a hardware-software (HW/SW) co-design methodology in which a CNN accelerator is described using high-level synthesis (HLS) tools that ease the parameterization of the design, facilitating more effective optimizations across multiple design constraints. Our experimental results demonstrate that the proposed design methodology is able to outperform non-parameterized design approaches, and it can be easily extended to other types of DL applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_04044 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | A Parameterizable Convolution Accelerator for Embedded Deep Learning Applications Mousouliotis, Panagiotis Keramidas, Georgios Computer Vision and Pattern Recognition Hardware Architecture Convolutional neural network (CNN) accelerators implemented on Field-Programmable Gate Arrays (FPGAs) are typically designed with a primary focus on maximizing performance, often measured in giga-operations per second (GOPS). However, real-life embedded deep learning (DL) applications impose multiple constraints related to latency, power consumption, area, and cost. This work presents a hardware-software (HW/SW) co-design methodology in which a CNN accelerator is described using high-level synthesis (HLS) tools that ease the parameterization of the design, facilitating more effective optimizations across multiple design constraints. Our experimental results demonstrate that the proposed design methodology is able to outperform non-parameterized design approaches, and it can be easily extended to other types of DL applications. |
| title | A Parameterizable Convolution Accelerator for Embedded Deep Learning Applications |
| topic | Computer Vision and Pattern Recognition Hardware Architecture |
| url | https://arxiv.org/abs/2602.04044 |