Orchestrating Multiple Mixed Precision Models on a Shared Precision-Scalable NPU
Open Access
- 20 June 2024
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 72-82
Abstract
No abstract availableFunding Information
- NRF (2022R1C1C1008131)
- Institute for Information and Communications Technology Promotion (2021-0-00310, 2020-0-01361, RS-2023-00277060)
This publication has 23 references indexed in Scilit:
- Efficient GPU-accelerated Join Optimization for Complex Queries2022
- Dataflow Mirroring: Architectural Support for Highly Efficient Fine-Grained Spatial Multitasking on Systolic-Array NPUs2021
- Gemmini: Enabling Systematic Deep-Learning Architecture Evaluation via Full-Stack Integration2021
- MLIR: Scaling Compiler Infrastructure for Domain Specific Computation2021
- Heterogeneous Dataflow Accelerators for Multi-DNN Workloads2021
- Layerweaver: Maximizing Resource Utilization of Neural Processing Units via Layer-Wise Scheduling2021
- PREMA: A Predictive Multi-Task Scheduling Algorithm For Preemptible Neural Processing Units2020
- HAWQ: Hessian AWare Quantization of Neural Networks With Mixed-Precision2019
- Acyclic Partitioning of Large Directed Acyclic Graphs2017
- An Efficient Parallelization Strategy for Dynamic Programming on GPU2013