A Resource-Driven Approach for Implementing CNNs on FPGAs Using Adaptive IPs
By Philippe Magalhães 1, Virginie Fresse 1, Benoît Suffran 2, Olivier Alata 1
1 Hubert Curien Laboratory - Université Jean Monnet
2 ST Microelectronics
Abstract
The increasing demand for real-time, low-latency artificial intelligence applications has propelled the use of Field-Programmable Gate Arrays (FPGAs) for Convolutional Neural Network (CNN) implementations. FPGAs offer reconfigurability, energy efficiency, and performance advantages over GPUs, making them suitable for edge devices and embedded systems. This work presents a novel library of resource-efficient convolution IPs designed to automatically adapt to the available FPGA resources. Developed in VHDL, these IPs are parameterizable and utilize fixed-point arithmetic for optimal performance. Four IPs are introduced, each tailored to specific resource constraints, offering flexibility in DSP usage, logic consumption, and precision. Experimental results on a Zynq UltraScale+ FPGA highlight the trade-offs between performance and resource usage. The comparison with recent FPGA-based CNN acceleration techniques emphasizes the versatility and independence of this approach from specific FPGA architectures or technological advancements. Future work will expand the library to include pooling and activation functions, enabling broader applicability and integration into CNN frameworks.
Index Terms—FPGA, CNN, Optimization, Adaptation
To read the full article, click here
Related Semiconductor IP
- CNN AI IP Core
- AI Accelerator Specifically for CNN
- Convolutional Neural Network (CNN) Compact Accelerator
Related White Papers
- An MDE Approach For Implementing Partial Dynamic Reconfiguration In FPGAs
- Implementing LTE on FPGAs
- Using an interface wrapper module to simplify implementing PCIe on FPGAs
- Implementing digital processing for automotive radar using SoC FPGAs
Latest White Papers
- SPAD: Specialized Prefill and Decode Hardware for Disaggregated LLM Inference
- DRsam: Detection of Fault-Based Microarchitectural Side-Channel Attacks in RISC-V Using Statistical Preprocessing and Association Rule Mining
- ShuffleV: A Microarchitectural Defense Strategy against Electromagnetic Side-Channel Attacks in Microprocessors
- Practical Considerations of LDPC Decoder Design in Communications Systems
- A Direct Memory Access Controller (DMAC) for Irregular Data Transfers on RISC-V Linux Systems