US 11,740,942 B2
Smart deployment of industrial IoT workloads
Lomash Kumar, Redmond, WA (US); Muhammad Usman Anwer, Seattle, WA (US); Nicolas Pouyez, Bainbridge Island, WA (US); Yazan Damiri, Seattle, WA (US); Rahul Nambiar, Vancouver (CA); and Glenn Danthi, Seattle, WA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Dec. 9, 2020, as Appl. No. 17/116,933.
Prior Publication US 2022/0179697 A1, Jun. 9, 2022
Int. Cl. G06F 9/50 (2006.01)
CPC G06F 9/505 (2013.01) [G06F 9/5044 (2013.01); G06F 9/5072 (2013.01); G06F 9/5083 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system, comprising:
one or more processors; and
one or more memories, wherein the one or more memories have stored thereon instructions, which when executed by the one or more processors, cause the one or more processors to implement a workload adaptation service for a plurality of clients, wherein the workload adaptation service is configured to, for individual clients:
receive, based on user input, an indication of a workload to be performed and at least one constraint for performance of the workload, wherein the at least one constraint comprises a maximum amount of time for output of a first portion of the workload to be received by a second portion of the workload;
determine a deployment model for the workload, wherein the determination takes into account the workload, one or more available client resources of a remote network of the client, one or more available provider resources of the provider network, and the at least one constraint for performance of the workload, and wherein the deployment model indicates:
an assignment of one or more portions of the workload for deployment to one or more of the client resources; and
an assignment of one or more other portions of the workload for deployment to one or more of the provider resources,
wherein the first portion and the second portion of the workload are assigned for deployment to one of the client resources or to one of the provider resources based on the maximum amount of time for the output of the first portion of the workload to be received by the second portion of the workload; and
deploy the one or more portions of the workload to the one or more client resources and the one or more other portions of the workload to the one or more provider resources.