RAELLA: Reforming the Arithmetic for Efficient, Low-Resolution, and Low-Loss Analog PIM: No Retraining Required!
Tanner Andrulis, Joel Emer, Vivienne Sze
Abstract
Processing-In-Memory (PIM) accelerators have the potential to efficiently run Deep Neural Network (DNN) inference by reducing costly data movement and by using resistive RAM (ReRAM) for efficient analog compute. Unfortunately, overall PIM accelerator efficiency is limited by energy-intensive analog-to-digital converters (ADCs). Furthermore, existing accelerators that reduce ADC cost do so by changing DNN weights or by using low-resolution ADCs that reduce output fidelity. These strategies harm DNN accuracy and/or require costly DNN retraining to compensate.
Topics & Concepts
Computer scienceConvertersRetrainingEfficient energy useArtificial neural networkInferenceArtificial intelligenceResistive random-access memoryFidelityFLOPSComputer engineeringEngineeringElectrical engineeringParallel computingTelecommunicationsVoltageBusinessInternational tradeAdvanced Memory and Neural ComputingCCD and CMOS Imaging SensorsFerroelectric and Negative Capacitance Devices