Layerweaver: Maximizing Resource Utilization of Neural Processing Units via Layer-Wise Scheduling.
Young H. OhSeonghak KimYunho JinSam SonJonghyun BaeJongsung LeeYeonhong ParkDong Uk KimTae Jun HamJae W. LeePublished in: HPCA (2021)
Keyphrases
- resource utilization
- processing units
- load balancing
- resource management
- quality of service
- parallel processing
- parallel computing
- response time
- computing systems
- high availability
- network resources
- energy consumption
- virtual machine
- video streaming
- grid computing
- computer systems
- multiple types
- power consumption
- user behavior
- frame rate
- network management
- data processing
- real time