Simultaneous Multikernel GPU: Multi-tasking throughput processors via fine-grained sharing
Author:
Publisher
IEEE
Link
http://xplorestaging.ieee.org/ielx7/7440961/7446041/7446078.pdf?arnumber=7446078
Cited by 107 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
1. ElasticRoom: Multi-Tenant DNN Inference Engine via Co-design with Resource-constrained Compilation and Strong Priority Scheduling;Proceedings of the 33rd International Symposium on High-Performance Parallel and Distributed Computing;2024-06-03
2. Towards Deterministic End-to-end Latency for Medical AI Systems in NVIDIA Holoscan;2024 ACM/IEEE 15th International Conference on Cyber-Physical Systems (ICCPS);2024-05-13
3. An Analysis of Collocation on GPUs for Deep Learning Training;Proceedings of the 4th Workshop on Machine Learning and Systems;2024-04-22
4. POSTER: FineCo: Fine-grained Heterogeneous Resource Management for Concurrent DNN Inferences;Proceedings of the 29th ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming;2024-02-20
5. MSHGN: Multi-scenario adaptive hierarchical spatial graph convolution network for GPU utilization prediction in heterogeneous GPU clusters;Journal of Parallel and Distributed Computing;2024-02
1.学者识别学者识别
2.学术分析学术分析
3.人才评估人才评估
"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370
www.globalauthorid.com
TOP
Copyright © 2019-2024 北京同舟云网络信息技术有限公司 京公网安备11010802033243号 京ICP备18003416号-3