English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Running multiple deep neural networks (DNNs) in parallel has become anemerging workload in both edge devices, such as mobile phones where multipletasks serve a single user for daily activities, and data centers, where variousrequests are raised from millions of users, as seen with large language models.To reduce the costly computational and memory requirements of these workloads,various efficient sparsification approaches have been introduced, resulting inwidespread sparsity across different types of DNN models. In this context,there is an emerging need for scheduling sparse multi-DNN workloads, a problemthat is largely unexplored in previous literature. This paper systematicallyanalyses the use-cases of multiple sparse DNNs and investigates theopportunities for optimizations. Based on these findings, we propose Dysta, anovel bi-level dynamic and static scheduler that utilizes both static sparsitypatterns and dynamic sparsity information for the sparse multi-DNN scheduling.Both static and dynamic components of Dysta are jointly designed at thesoftware and hardware levels, respectively, to improve and refine thescheduling approach. To facilitate future progress in the study of this classof workloads, we construct a public benchmark that contains sparse multi-DNNworkloads across different deployment scenarios, spanning from mobile phonesand AR/VR wearables to data centers. A comprehensive evaluation on the sparsemulti-DNN benchmark demonstrates that our proposed approach outperforms thestate-of-the-art methods with up to 10% decrease in latency constraintviolation rate and nearly 4X reduction in average normalized turnaround time.Our artifacts and code are publicly available at:https://github.com/SamsungLabs/Sparse-Multi-DNN-Scheduling.

Sparse-DySta: Sparsity-Aware Dynamic and Static Scheduling for Sparse Multi-DNN Workloads