‘It’s about basically taking our Nutanix Cloud Platform, our Kubernetes Platform, and our AMD Enterprise AI software, getting them all to work on top of AMD accelerated compute hardware. So servers using AMD GPUs effectively, and bringing that solution to market. As part of that, AMD is doing two things. They’re investing in buying $150 million worth of Nutanix common stock. The second thing they’re doing is they’re funding us up to $100 million in a combination of both R&D to bring the platform to market, and go-to-market efforts,’ says Nutanix CEO Rajiv Ramaswami.
AMD and Nutanix Wednesday said they have signed a multi-year strategic partnership aimed at developing what they called an open, full-stack AI infrastructure platform specifically for agentic AI applications.
As part of the agreement, AMD will purchase $150 million in Nutanix common stock at $36.26 per share. AMD will also fund up to $100 million for Nutanix to help with joint engineering initiatives and go-to-market collaboration.
The move comes as Nutanix evolves into what it calls a platform company. Nutanix CEO Rajiv Ramaswami in August told CRN his company wanted to become a “de facto platform” that would let customers run applications and manage data regardless of where it resides.
[Related: Nutanix Takes ‘Cloud To The Next Level’ With New 7.5 Platform, CEO Says]
AMD was not able to provide CRN further information by press time.
Ramaswami, in a press conference after the company’s quarterly financial conference call, said the new strategic partnership with AMD is an extension of an existing partnership between the two. As part of that partnership, the two companies will work to get the Nutanix Cloud Platform, the Nutanix Kubernetes Platform, and the AMD Enterprise AI software to work on top of AMD accelerated compute hardware.
“So servers using AMD GPUs effectively, and bringing that solution to market,” he said. “As part of that, AMD is doing two things. They’re investing in buying $150 million worth of Nutanix common stock. The second thing they’re doing is they’re funding us up to $100 million in a combination of both R&D to bring the platform to market, and go-to-market efforts.”
Nutanix previously had an AMD relationship, Ramaswami said.
“We already support Intel servers and AMD servers,” he said. “That’s been there for a while. Nothing new on that front. And on that front, all we’re doing is we continue to expand the repertoire of AMD chipsets that we support from a server perspective. A lot of the focus on this expanded partnership around AMD GPUs accelerated compute.
So historically, we have worked with Nvidia GPUs and taken that solution to market. We’ve had that in the market for a while, but now we’re bringing on AMD solutions.”
When asked what Nutanix and AMD plan to use their jointly-developed platform for, Ramaswami said enterprises are building AI applications.
“They are consumers of AI,” he said. “They are building inferencing applications. They’re building agentic AI applications. Very early stage on the agentic side, but people are already doing it. For example, we internally at Nutanix have a customer support application. We have other internal applications around document search and so forth that we built, taking open-source models, running them on top of the Nutanix stack on top of Nvidia GPUs. What we’re doing with this platform is we are providing the infrastructure software along with the hardware from AMD through server partners for companies to run their enterprise inferencing and agentic AI applications and manage all the data associated with that.”
Nutanix offers technology platforms built with a variety of technology partners, Ramaswami said.
“Our goal is to provide customer choice and bring to our customers a variety of underlying platforms,” he said. “There’s multiple underlying hardware platforms. Nvidia of course has been the market leader for quite a while here in the space. AMD is the other big platform company when it comes to GPUs. Customers want choice across both. You’ve seen, for example, how hard it has been to get ahold of Nvidia GPUs, for example, and then there’s pricing comparisons and so forth. So there’s a lot of reasons why it’s good to have multiple providers in the market. And from a Nutanix perspective, we want to be able to support all the major providers and provide choice to our customers.”
When asked about potential supply issues related to AMD GPUs, Ramaswami said there are supply issues everywhere now.
“There’s obviously memory shortages, there’s CPU shortages, and there’s some GPU shortages as well,” he said. “There’s shortages of everything. But that, to me, was not a driver for us doing anything with AMD. This was much more of a strategic partnership to really bring a complete solution to the market for enterprises and service providers.”
Nutanix said the partnership will optimize the Nutanix Cloud and Nutanix Kubernetes Platforms on AMD EPYC CPUs and AMD Instinct GPUs. The two will also integrate the AMD ROCm software ecosystem and the AMD Enterprise AI platform into Nutanix AI full-stack solutions to build an open solution for agentic AI platforms that will be supported by multiple OEM partners.
The first agentic AI platform developed jointly by Nutanix and AMD is slated be available some late calendar 2026.
The jointly developed platform will also support high-performance AI inference accelerations powered by AMD Instinct GPUs and EPYC CPUs, high-core-density compute and orchestration through AMD EPYC processors, and unified lifecycle management via Nutanix Enterprise AI, Nutanix said.
