mlsanityData to 7 Oct 2026

Fast Convolutional Nets With fbfft: A GPU Performance Evaluation

arXiv 1412.7580 ↗First posted 24 Dec 2014cs.LG cs.DC cs.NEcited by 27

Not on the map. Why?

Abstract (arXiv)

We examine the performance profile of Convolutional Neural Network training on the current generation of NVIDIA Graphics Processing Units. We introduce two new Fast Fourier Transform convolution implementations: one based on NVIDIA's cuFFT library, and another based on a Facebook authored FFT implementation, fbfft, that provides significant speedups over cuFFT (over 1.5x) for whole CNNs. Both of these convolution implementations are available in open source, and are faster than NVIDIA's cuDNN implementation for many common convolutional layers (up to 23.5x for some synthetic kernel configurations). We discuss different performance regimes of convolutions, comparing areas where straightforward time domain convolutions outperform Fourier frequency domain convolutions. Details on algorithmic applications of NVIDIA GPU hardware specifics in the implementation of fbfft are also provided.

Key facts. Cited by 27 papers in the map (data through 7 Oct 2026). First posted on arXiv 24 Dec 2014.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 9; 2022: 5; 2023: 3; 2024: 4; 2025: 4; 2026: 2.

27 papers.

Sort:

  1. 2021Scaling Local Self-Attention for Parameter Efficient Visual Backbonescited by 220
  2. 2021ReduNet: A White-box Deep Network from the Principle of Maximizing Rate Reductioncited by 84
  3. 2021Manipulating Identical Filter Redundancy for Efficient Pruning on Deep and Complicated CNNcited by 9
  4. 2021Resource-Efficient Deep Learning: A Survey on Model-, Arithmetic-, and Implementation-Level Techniquescited by 9
  5. 2023As large as it gets: Learning infinitely large Filters via Neural Implicit Functions in the Fourier Domaincited by 8
  6. 2021Knowledge Distillation Circumvents Nonlinearity for Optical Convolutional Neural Networkscited by 6
  7. 2022Fast-FNet: Accelerating Transformer Encoder Models via Efficient Fourier Layerscited by 5
  8. 2023AdaOPC: A Self-Adaptive Mask Optimization Framework For Real Design Patternscited by 5
  9. 2025Model Compression using Progressive Channel Pruningcited by 4
  10. 2021FourierNets enable the design of highly non-local optical encoders for computational imagingcited by 3
  11. 2024From Algorithm to Hardware: A Survey on Efficient and Safe Deployment of Deep Neural Networkscited by 2
  12. 2021cuConv: A CUDA Implementation of Convolution for CNN Inferencecited by 1
15 more
  1. 2021Efficient and Generic 1D Dilated Convolution Layer for Deep Learningcited by 1
  2. 2022Learning Convolutional Neural Networks in the Frequency Domaincited by 1
  3. 2022Kernel-Segregated Transpose Convolution Operationcited by 1
  4. 2023Low-Rank Winograd Transformation for 3D Convolutional Neural Networkscited by 1
  5. 2021Convolutional Neural Network Simplification with Progressive Retrainingcited by 0
  6. 2022Testing predictions of representation cost theory with CNNscited by 0
  7. 2022Pruning Very Deep Neural Network Channels for Efficient Inferencecited by 0
  8. 2024TFDMNet: A Novel Network Structure Combines the Time Domain and Frequency Domain Featurescited by 0
  9. 2024Phasor-Driven Acceleration for FFT-based CNNscited by 0
  10. 2024Flexible Coded Distributed Convolution Computing for Enhanced Straggler Resilience and Numerical Stability in Distributed CNNscited by 0
  11. 2025Underwater object detection in sonar imagery with detection transformer and Zero-shot neural architecture searchcited by 0
  12. 2025S3F-Net: A Multi-Modal Approach to Medical Image Classification via Spatial-Spectral Summarizer Fusion Networkcited by 0
  13. 2025Accelerated Rotation-Invariant Convolution for UAV Image Segmentationcited by 0
  14. 2026Spectral-Structured Diffusion for Single-Image Rain Removalcited by 0
  15. 2026Flash EQ-Linear: Accelerating Equivariant Linear Layers via Group-wise Discrete Fourier Transformcited by 0