Dell touts VxRail and GPU combo for Nvidia’s AI Enterprise suite

Dell has announced that its VxRail hyperconverged systems are certified for Nvidia’s newly available AI Enterprise software suite, making it simpler for organisations to deploy GPU-accelerated infrastructure for processing AI and data analytics workloads.

VxRail is based on Dell EMC PowerEdge server hardware and is designed to provide a fully integrated and pre-configured HCI system optimised for VMware’s hypervisor with vSAN providing the storage layer. This has made VxRail a popular choice for enterprises that wish to stand up virtualised infrastructure with a minimum of fuss.

Meanwhile, Nvidia’s AI Enterprise is described as a comprehensive software suite of AI tools and frameworks that has likewise been optimised for systems running VMware vSphere and certified by Nvidia. Unsurprisingly, Dell’s VxRail is the first HCI system to achieve such certification.

In a blog posting, Dell’s senior manager for CI/HCI product marketing, Nancy Hurley, stated that the VxRail certified systems have been tested specifically with Nvidia AI Enterprise workloads.

“While VxRail and Nvidia GPUs deliver the performance needed for AI workloads, the big story here is how we have worked together to simplify the entire experience — from procurement to deployment, to day to day operations and lifecycle management,” she said.

VxRail configurations that have been tested with Nvidia AI Enterprise comprise the V570, V570F and V670F, and choosing these should enable customers to deploy faster, since these offer known working configurations for the software stack.

In June, Dell announced that the VxRail V Series is available with Nvidia A100 Tensor Core GPU options. When combined with Nvidia AI Enterprise and NVMe caching capabilities, these configurations offer greater performance and simpler deployment for demanding AI and machine learning applications, the firm claims.

Nvidia’s AI Enterprise suite includes the RAPIDS software development kit, which supports accelerated data science on GPUs, plus Nvidia’s Triton Inference Server which simplifies the deployment of AI models at scale and machine learning frameworks like TensorFlow and PyTorch.