TY - GEN
T1 - Workload Intelligence
T2 - 2025 International Conference for High Performance Computing, Networking, Storage, and Analysis, SC 2025
AU - Huang, Lexiang
AU - Parayil, Anjaly
AU - Zhang, Jue
AU - Qin, Xiaoting
AU - Bansal, Chetan
AU - Stojkovic, Jovan
AU - Zardoshti, Pantea
AU - Misra, Pulkit
AU - Cortez, Eli
AU - Ghelman, Raphael
AU - Goiri, Íñigo
AU - Rajmohan, Saravan
AU - Kleewein, Jim
AU - Fonseca, Rodrigo
AU - Zhu, Timothy
AU - Bianchini, Ricardo
N1 - Publisher Copyright:
© 2025 Copyright held by the owner/author(s).
PY - 2025/11/15
Y1 - 2025/11/15
N2 - Today, cloud workloads are largely opaque to the cloud platform. Typically, the only information the platform receives is the virtual machine (VM) type and possibly a decoration to the type (e.g., the VM is evictable). Similarly, workloads receive minimal information from the platform; generally, only telemetry from their VMs or occasional signals (e.g., just before a VM is evicted). The narrow interface between workloads and platforms has several drawbacks: (1) a surge in VM types and decorations in public cloud platforms complicates customer selection; (2) key workload characteristics (e.g., low availability requirements) are often unspecified, hindering platform customization for optimized resource usage and cost savings; and (3) workloads may be unaware of potential optimizations or lack sufficient time to react to platform events. To resolve these issues and improve cloud efficiency, we propose Workload Intelligence (WI), a framework for enabling dynamic bi-directional communication between cloud workloads and cloud platform.
AB - Today, cloud workloads are largely opaque to the cloud platform. Typically, the only information the platform receives is the virtual machine (VM) type and possibly a decoration to the type (e.g., the VM is evictable). Similarly, workloads receive minimal information from the platform; generally, only telemetry from their VMs or occasional signals (e.g., just before a VM is evicted). The narrow interface between workloads and platforms has several drawbacks: (1) a surge in VM types and decorations in public cloud platforms complicates customer selection; (2) key workload characteristics (e.g., low availability requirements) are often unspecified, hindering platform customization for optimized resource usage and cost savings; and (3) workloads may be unaware of potential optimizations or lack sufficient time to react to platform events. To resolve these issues and improve cloud efficiency, we propose Workload Intelligence (WI), a framework for enabling dynamic bi-directional communication between cloud workloads and cloud platform.
UR - https://www.scopus.com/pages/publications/105023970946
UR - https://www.scopus.com/pages/publications/105023970946#tab=citedBy
U2 - 10.1145/3712285.3759848
DO - 10.1145/3712285.3759848
M3 - Conference contribution
AN - SCOPUS:105023970946
T3 - Proceedings of the International Conference for High Performance Computing, Networking, Storage, and Analysis, SC 2025
SP - 2203
EP - 2215
BT - Proceedings of the International Conference for High Performance Computing, Networking, Storage, and Analysis, SC 2025
PB - Association for Computing Machinery, Inc
Y2 - 16 November 2025 through 21 November 2025
ER -