VMware Powers AI/ML Applications via GPU Sharing
Virtualization kingpin VMware is updating its flagship VMware VSphere 7 offering, providing the ability to provision elastic infrastructure for artificial intelligence (AI) and machine learning (ML) applications.
By "elastic," the company means GPU hardware-acceleration horsepower can now be pooled and shared across the network where needed for compute-intensive AI/ML applications.
VMware leveraged technology gained from its acquisition of Bitfusion to power the new service.
It's expected to roll out soon to VMware VSphere 7, which provides virtualization technology with a focus on hybrid cloud computing and which debuted earlier this year.
The new tech is called VMware vSphere Bitfusion, reflecting its roots in Bitfusion, acquired by VMware last year. That company was described as a pioneer in the virtualization of hardware accelerated devices, focusing on GPU technology.
The new hardware acceleration feature will boost AI/ML workload performance by enabling organizations to pool GPU resources on servers -- rather than being isolated on specific servers -- and then share the computing power within datacenters as a pool of network-accessible resources that can be used in virtual machines or containers.
"Organizations use hardware accelerators such as GPUs to dramatically improve the performance of AI/ML workloads that may run several hours or longer," VMware said in a June 2 news release. "IT teams have come to realize that these hardware accelerators are isolated islands -- unable to be shared across many parts of the business. The inability to share those resources leads to inefficient and poor utilization of both existing and newly purchased resources. The combination of Bitfusion and VMware vSphere will help organizations achieve cost savings, enable resource sharing out of the box, and deliver the right hardware accelerator resource, like a GPU, to the right workload at the right time."
A blog post provides more information: "vSphere Bitfusion delivers elastic infrastructure for AI/ML workloads by creating pools of hardware accelerator resources. The best-known accelerators today are GPUs which vSphere can now use to create AI/ML cloud pools that can be used on-demand. GPUs can now be used efficiently across the network and driven to the highest levels of utilization possible. This means it allows for the sharing of GPUs in a similar fashion to the way vSphere allowed the sharing of CPUs many years ago. The result is an end to isolated islands of inefficiently used resources. End-users and service providers (wanting to off GPU as a Service for example) are going to see big benefits with this new feature."
As part of the VMware vSphere Enterprise Plus edition, VMware vSphere Bitfusion is expected to be generally available in the second quarter of VMware's fiscal year 2021, which runs from May 2, 2020 to July 31, 2020.
David Ramel is an editor and writer for Converge360.