Nvidia revealed in an announcement yesterday that VMware was an official business partner with its new updates to its virtual GPU technology. Namely, a hybrid cloud service on AWS that provides online access to high-end number-crunching power for modern integrated AI applications.
VMware Cloud will be the primary service to avail Nvidia’s upgrades. As per its original technical design, it will be able to allow easy migration of virtualization apps, which would now have the option for accelerated high-speed computing for intensive applications such as AI training/machine learning and various data science workflows.
vCompute Server was also another announced service that benefits from this cloud-management plus vGPU partnership. This will provide the platform to directly offload processes and operations in virtualized workstations and/or offices, empowering VMware’s main service suite, which includes vSphere, vCenter, vMotion, and of course, WMware Cloud.
Additionally, offloading vCompute Server workflows can be easily transferred to the cloud in just a few option tweaks. This provides yet another option to seamless connect work environments, as well as provide universal access to data whether it is automated for a system, or specifically requested by a user.
In accordance with all of this, VMware Cloud on AWS will also offer EC2 instances, thanks to Nvidia’s enterprise-level T4 100 GPUs. This additional service upgrade is slated to be implemented in March next year.
While most of these new upgraded services are advertised for heavy AI applications such as deep learning and machine learning, Nvidia actually presents this also as a way to further streamline virtual office operations focused in 3D design. Like for example when editing a video, or when working on a schematic project with a team of other designers collaborating remotely.
Nvidia and VMware’s “GPU-accelerated hybrid cloud infrastructure” is slated to promote innovation within the industry that primarily uses such services. Indeed, when you no longer have to worry about hardware requirements for heavy data workflows, or get anxious about making accumulated data universally compatible, new applications combining both company’s offerings might soon be brought to life.