Nvidia Corp. wants to tighten its grip on the enterprise artificial intelligence market, and it’s looking to do so by strengthening its partnership with Lenovo Group Ltd.

In a new collaboration announced today, it said it’s working with Lenovo to deliver advanced, hybrid computing platforms capable of powering generative AI workloads in every location where data is created.

Nvidia Chief Executive Jensen Huang announced the new offerings alongside Lenovo Chairman and CEO Yuanqing Yang during a keynote at the annual Lenovo Tech World conference in Austin today. Onstage, he discussed the growing demand from businesses for accelerated computing systems, AI software and expert services as they seek to build and run custom AI models trained on their own data.

To cater to this demand, Lenovo announced new, on-premises systems that are optimized to run the Nvidia AI Enterprise software stack and power AI workloads anywhere. They include the new Lenovo ThinkSystem SR675 V3 server and the ThinkStation PX workstation, which are available now. With the Nvidia NeMoTM framework that’s included in Nvidia AI Enterprise, businesses can now customize a selection of large language models found within Nvidia AI Foundations, the companies said. By using the latest retrieval-augmented generation techniques and fine-tuning methods, organizations can build generative AI applications trained on their own data to handle specific business use cases.

The ThinkSystem SR675V3 combines Nvidia’s L40S graphics processing units with BlueField-3 data processing units and Spectrum-X networking to power AI workloads in any data center.

As for the ThinkStation PX, it brings expanded AI capabilities and data center-like performance to any location, including the network edge, with up to four Nvidia RTX 600 Ada GPUs on board.

In addition to the hardware, Nvidia and Lenovo will collaborate to build even more powerful systems based on the flexible Nvidia MGX modular reference design. These systems will be built to handle the most demanding of generative AI workloads, they said. Customers will be able to combine this hardware with the Nvidia Omniverse platform to implement immersive simulations and cognitive decisions at enterprise scale.

Bob Pette, Nvidia’s vice president of enterprise platforms, said enterprises need solutions that give them the flexibility to develop and deploy AI on workstations, data centers and the cloud. “Nvidia and Lenovo are providing a comprehensive portfolio of hybrid AI systems that businesses can rely on to power generative AI from virtually anywhere,” he added.

Lenovo’s systems will also support the recently announced VMware Private AI Foundation with Nvidia, paving the way for thousands of VMware customers to implement generative AI.

Finally, Lenovo announced the creation of a new AI Professional Services Practice and TruScale software-as-a-service model that allows businesses to access its hardware on a pay-as-you-go model, thereby reducing the initial investment required to get generative AI up and running.

Lenovo’s AI Professional Services Practice is designed to help businesses of any size navigate the generative AI landscape, identify use cases and get new applications up and running in the most cost effective way, the company said.

Kirk Skaugen, president of Lenovo Infrastructure Solutions Group, said his company’s partnership with Nvidia will help companies to unlock the power of their data anywhere it is created. “Lenovo’s expanded engineering and time to market partnership with Nvidia simplifies the path to generative AI for all and helps customers anywhere rapidly use cutting-edge AI to transform their business.”

Image: Nvidia