Low Precision Floating Point Arithmetic for High Performance FPGA-based CNN Acceleration
Chen Wu, Mingyu Wang, Xinyuan Chu, Kun Wang, Lei He
Abstract
Low precision data representation is important to reduce storage size and memory access for convolutional neural networks (CNNs). Yet, existing methods have two major limitations: (1) requiring re-training to maintain accuracy for deep CNNs, and (2) needing 16-bit floating point or 8-bit fixed point for a good accuracy.
Topics & Concepts
Convolutional neural networkComputer scienceFloating pointField-programmable gate arrayFixed-point arithmeticPoint (geometry)Representation (politics)Double-precision floating-point formatAccelerationSingle-precision floating-point formatHardware accelerationAlgorithmComputer hardwareArithmeticArtificial intelligenceParallel computingMathematicsGeometryPolitical scienceClassical mechanicsLawPoliticsPhysicsNeural Networks and ApplicationsNeural Networks and Reservoir ComputingModel Reduction and Neural Networks