Interest in AI Workloads has led major enterprise vendors to cater to the needs of organizations by providing packaged solutions. Nutanix is the latest vendor to offer an AI-focused solution called “GPT-in-a-Box.” This offering aims to address concerns related to cost, data governance, and return on investment. Many organizations are interested in AI but are hesitant to experiment in the cloud due to potential high costs and data compliance requirements.

Nutanix’s solution includes its software stack, which abstracts hardware hyperscale clouds into hybrid clouds and automates management. GPT-in-a-Box is specifically designed for AI workloads and incorporates Kubeflow for easy deployment of machine learning workloads. The PyTorch machine learning framework and Jupyter are also included. Users can run their preferred generative pre-trained transformers (GPTs), with Llama2, Falcon, and mosaicML supported.

Nutanix’s AHV hypervisor is already certified to run Nvidia AI Enterprise 4.0, a key AI workload, and can also manage GPUs. The company recognizes that AI workloads require substantial amounts of data, which may be best stored in the cloud while running on-premises or at the edge. Nutanix’s platform is designed to manage these workloads in a logical construct.

Nutanix aims to encourage customers already using its platform to consider it for their AI workloads instead of creating separate silos. Other major vendors, such as Dell and VMware, have also introduced similar offerings. The big hyperscale clouds have already packaged AI workloads as well.

Currently, Nutanix supports Nvidia GPUs, and the company plans to support AMD’s accelerators in the future. Support for Intel’s standalone GPUs is not yet considered, but Nutanix does not rule out the possibility in the future.

The post Nutanix Launches “GPT-in-a-Box Solution” for AI Workloads appeared first on TS2 SPACE.