9 January 2026 in Journal of Systems Architecture
11 February 2026 in IEEE Embedded Systems Letters
GCN accelerators often adopt pruning or quantization to reduce resource demands, yet energy efficiency remains a key challenge on FPGA platforms due to DSP-based matrix multiplications. Existing DSP-free approximate multipliers fail to strike a good trade-off between precision and efficiency. To overcome these, this letter presents the first FPGA-based FP8 approximate GCN accelerator, enabling both low precision loss and high energy efficiency. The proposed architecture is deployed on a Kintex-7 K325T FPGA, demonstrating that the accuracy loss of the approximate FP8 GCN model remains within 3%. While achieving average performance improvements of 12.9× (0.7-44.0×) and average energy savings of 169× (1.01-1010×) over CPUs, GPUs, and prior edge-FPGA-based accelerators.
6 March 2026 in Integration
4 May 2026 in IEEE Transactions on Very Large Scale Integration (VLSI) Systems
The 8-bit floating-point (FP8) format has gained growing interest in neural networks (NNs) for its superior dynamic range over traditional INT8. However, multiply-accumulate (MAC) operations remain a major source of power consumption during inference of NNs, which makes DSP-free design important, especially for edge FPGAs with few or no DSPs. Therefore, this brief presents FPGA-based FP8 approximate neural network engine (FANE), an FPGA-based approximate NN engine for FP8. We first introduce a novel approximation method that replaces the multiplications by linear additions. This approximate method reduces power consumption while maintaining high accuracy, outperforming the latest FP8 approximate multiplier by 53.15%. Based on this design, we construct an FP8 MAC unit and integrate it into both a convolution engine and a matrix–vector multiplication (MVM) unit. Finally, we integrate our design into a large language model (LLM). The result shows 61.5% higher efficiency (TOPS/W) than the previous design, demonstrating the superiority of FANE in terms of performance and power efficiency. The code of FANE is available on our https://github.com/hanbao04/FANE-FPGA-based-FP8-Approximate-Neural-Network-Engine.git