A 28nm 1.644TFLOPS/W Floating-Point Computation SRAM Macro with Variable Precision for Deep Neural Network Inference and Training
Sangsu Jeong, Jeongwoo Park, Dongsuk Jeon
Abstract
This paper presents a digital compute-in-memory (CIM) macro for accelerating deep neural networks. The macro provides high-precision computation required for training deep neural networks and running state-of-the-art models by supporting floating-point MAC operations. Additionally, the design supports variable computation precision, enabling optimized processing for different models and tasks. The design achieves 1.644TFLOPS/W energy efficiency and 57.9GFLOPS/mm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> computation density while supporting a wide range of floating-point data formats and computation precisions.
Topics & Concepts
ComputationComputer scienceFloating pointArtificial neural networkMacroStatic random-access memoryInferenceVariable (mathematics)Computer engineeringPoint (geometry)Artificial intelligenceRange (aeronautics)Parallel computingAlgorithmComputer hardwareMathematicsEngineeringProgramming languageGeometryAerospace engineeringMathematical analysisAdvanced Memory and Neural ComputingFerroelectric and Negative Capacitance DevicesParallel Computing and Optimization Techniques