USENIX

When will my ML Job finish? Toward providing Completion Time Estimates through Predictability-Centric Scheduling

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are ...
USENIX

Looking Beyond GPUs for DNN Scheduling on Multi-Tenant Clusters

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are ...
USENIX

Themis: Fair and Efficient GPU Cluster Scheduling

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are ...
USENIX

CASSINI: Network-Aware Job Scheduling in Machine Learning Clusters

We present CASSINI, a network-aware job scheduler for machine learning (ML) clusters. CASSINI introduces a novel geometric abstraction to consider the communication pattern of different jobs while ...
USENIX

AntMan: Dynamic Scaling on GPU Clusters for Deep Learning

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are ...
USENIX

ATP: In-network Aggregation for Multi-tenant Learning

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are ...
USENIX

CarMap: Fast 3D Feature Map Updates for Automobiles

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are ...
USENIX

MQSim: A Framework for Enabling Realistic Studies of Modern Multi-Queue SSD Devices

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are ...
USENIX

SkyPilot: An Intercloud Broker for Sky Computing

To comply with the increasing number of government regulations about data placement and processing, and to protect themselves against major cloud outages, many users want the ability to easily migrate ...
USENIX

XRP: In-Kernel Storage Functions with eBPF

With the emergence of microsecond-scale NVMe storage devices, the Linux kernel storage stack overhead has become significant, almost doubling access times. We present XRP, a framework that allows ...
USENIX

Millions of Tiny Databases

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are ...
USENIX

OptiReduce: Resilient and Tail-Optimal AllReduce for Distributed Deep Learning in the Cloud

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are ...