Vultr, a leading independent cloud infrastructure provider, announced today that Vultr Talon, powered by NVIDIA GPUs and NVIDIA AI Enterprise software, is currently in beta. Vultr Talon, a ground-breaking cloud-based platform, enables GPU sharing, allowing several workloads to execute effectively on a single NVIDIA GPU.
Vultr is the first cloud service to provide GPU sharing by virtualizing NVIDIA A100 Tensor Core GPUs. Vultr is launching a set of virtual machine (VM) packages incorporating the NVIDIA A100 today, with prices beginning at $90 per month, or $0.13 per hour.
Breakthrough Cloud Platform Enabling AI Workloads at a Fraction of the Cost
Vultr Talon is powered by a cutting-edge NVIDIA GPU virtualization technology based on NVIDIA's vGPU software. Rather than connecting whole real GPUs to VMs, Vultr adds just a portion in the form of a virtual GPU (vGPU). The NVIDIA AI Enterprise software package, which incorporates NVIDIA vGPU software and is geared for remotely performing AI workloads and high-performance data analytics, powers these vGPUs.
A vGPU appears to a customer's system just like an actual GPU. Each vGPU has its own devoted memory, which is a fraction of the memory on the underlying card. The vGPU has access to a fraction of the computing capacity of the actual GPU. NVIDIA's Multi-Instance GPU (MIG) technology is utilized for Vultr plans with at least 10GB of GPU RAM to give tenants with assured QoS, completely isolated GPU high-bandwidth memory cache, and dedicated compute cores.
High-end GPUs can cost hundreds of dollars per month to rent. While this cost is often justified for the most demanding corporate workloads, for many enterprises and developers, the cost of even a single GPU might be prohibitive to getting started, experimenting, or running apps in development and testing environments. Even organizations with large IT budgets may wind up losing money by supplying more GPU capacity than required or just electing not to use GPUs at all.
At Vultr, we pride ourselves on making high-performance cloud infrastructure affordable for everyone. With Vultr Talon, we have turned the GPU delivery model upside down. Because of our breakthrough cloud platform, AI developers and data scientists can provision exactly as much NVIDIA GPU processing as they actually need, at prices they can afford.”
J.J. Kardwell, CEO of Vultr’s parent company,