English

Enabling Long FFT Convolutions on Memory-Constrained FPGAs via Chunking

Machine Learning 2026-01-13 v1 Hardware Architecture

Abstract

The need for long-context reasoning has led to alternative neural network architectures besides Transformers and self-attention, a popular model being Hyena, which employs causal 1D-convolutions implemented with FFTs. Long convolutions enable efficient global context mixing, but requirements for intermediate results exceed the 2-3 MB Block RAM capacity of FPGAs. We present a chunked FFT convolution approach enabling 450K length sequence by 450K length filter convolutions on an Alveo U200 FPGA with 2.8 MB BRAM through chunking and overlap-add reconstruction. We find that throughput scales proportionally with chunk size while degrading minimally by 7% for our longest sequences, demonstrating that careful memory management enables deployment of long-context primitives on edge FPGAs without sacrificing performance.

Keywords

Cite

@article{arxiv.2601.06065,
  title  = {Enabling Long FFT Convolutions on Memory-Constrained FPGAs via Chunking},
  author = {Peter Wang and Neelesh Gupta and Viktor Prasanna},
  journal= {arXiv preprint arXiv:2601.06065},
  year   = {2026}
}

Comments

2 pages, submitted to 2025 HiPC Conference

R2 v1 2026-07-01T08:58:09.033Z