trahko - stock.adobe.com

Nutanix CEO maps out agentic AI strategy, targets VMware defectors

Rajiv Ramaswami talks up Nutanix’s agentic AI play, the growing demand for sovereign cloud capabilities, and why decoupling storage from HCI is hastening migrations from VMware

Nutanix wants to break down enterprise infrastructure silos, pitching itself as the de facto platform for everything from legacy virtual machines (VMs) to modern Kubernetes containers and emergent AI agents, according to its CEO Rajiv Ramaswami

“Our view is that we want to be the platform for all applications, right, and wherever. I mean, in very simple terms, data everywhere,” he told Computer Weekly during a recent visit to Singapore.

The enterprise software provider recently rolled out a slew of major announcements, including bare-metal support for the Nutanix Kubernetes Platform and a new agentic AI platform. According to Ramaswami, the goal is to provide a modernisation pathway for IT leaders without deploying disparate infrastructure silos.

Nutanix’s move comes at a time when enterprises are advancing in their AI journey, from simple inferencing – like document search and summarisation – towards autonomous AI agents capable of  executing complex, multi-step tasks independently.

While open-source agentic tools are readily available for developers to download, Ramaswami warned that using them without guardrails puts organisations at risk. To address this, Nutanix has integrated an AI gateway into its infrastructure stack, giving CIOs granular control over who accesses models and how much computing power they consume.

“You can put limits on those,” Ramaswami said. “You can say this HR department is only going to be allowed to use this set of agents and this many tokens from this particular model. You can have that policy instituted, and you can put that in place and regulate it.”

Running models on on-premises infrastructure also bounds the financial risk. “When I talk to CIOs, one of the things they keep asking about is, ‘Can you get me away from paying on a per-token basis?’ The answer to that is, you build your own infrastructure, and then you control it. You know exactly what your cost is ... whatever you do fits within the cost envelope.”

To support the growing demand for AI infrastructure, Nutanix supports all Nvidia graphics processing units (GPUs), with plans to support AMD chips next. Ramaswami noted the company will also monitor market demand for custom accelerators such as Google’s tensor processing units (TPUs) and Groq’s language processing units (LPUs) before committing to further integrations.

Capitalising on the VMware exodus

Nutanix’s AI ambition coincides with a major infrastructure land grab. Following Broadcom’s acquisition of VMware, Nutanix has been aggressively courting customers looking for an alternative virtualisation provider.

According to Ramaswami, the company is seeing “about 500 to 1,000 new customers every quarter”, almost all of which are migrations from VMware.

To grease the wheels for these defectors, Nutanix made an uncharacteristic move for a company founded on hyperconverged infrastructure (HCI): it announced support for multiple external storage providers, including Dell, Pure Storage, and NetApp.

“It’s a much easier migration path for customers to, for example, leave Broadcom and come to us, if we can just work on their existing hardware,” Ramaswami said. He added that ongoing hardware supply chain shortages make buying net-new equipment an expensive and slow hurdle. “The fact that now they can use Nutanix with their existing deployed systems makes it easier for them to prioritise their projects and get it done.”

The rise of sovereign cloud

While in Singapore, Ramaswami also noted growing appetite for sovereign cloud offerings across the Asia-Pacific region.

“Sovereignty is about three things,” he explained. “One is localisation ... The second is self-reliance, meaning you want your own infrastructure operated by your own people and controlled by yourself. The third portion is data sovereignty to make sure data remains in-country.”

Ramaswami noted that recent global conflicts and major public cloud outages have served as a wake-up call for enterprises and governments. “It has opened people’s eyes, saying, ‘I can’t just assume that because it’s in the public cloud, somebody is going to take care of it for me. I need to worry about it myself and make sure that I have a resiliency plan in place’.”

To offer that resilience across modern workloads, Nutanix is expanding its container strategy. Noting that Kubernetes can’t always maximise hardware utilisation when layered over VMs, the company’s new bare-metal support for containers is aimed at both heavyweight and lightweight Kubernetes applications at the edge of the network.

“For us, we want to provide the maximum flexibility to customers. You can deploy your containers on bare metal, you can deploy them on top of VMs, you can deploy them on top of native public cloud ... and we can help you manage all of that,” Ramaswami said, adding that Nutanix provides the exact same network, security and database experience across all environments.

Financial maturity and the rule of 40

Ultimately, Nutanix expects its AI strategy of bridging today’s VMs with tomorrow’s containers and AI platforms to fuel long-term financial health.

Outlining the company’s 2029 investor roadmap, Ramaswami confidently projected mid-to-high teens revenue growth while continuing to deliver leverage on the bottom line with its operating income to maintain a “sustainable rule of 40”.

A widely used financial benchmark in the software industry, the rule of 40 dictates that a healthy tech company’s combined revenue growth rate and profit margin should equal or exceed 40%. It serves as a balancing scale for investors: if a company’s revenue growth slows, its profitability must increase to compensate.

By targeting mid-to-high teens revenue growth, Ramaswami is signalling to Wall Street that Nutanix aims to achieve and sustain operating profit margins of over 20%. For enterprise buyers and investors alike, Nutanix’s goal is to deliver steady, disciplined expansion while bringing healthy returns to the bottom line, proving it can fund its AI and infrastructure ambitions without burning through cash.

Read more about AI infrastructure in APAC

Read more on Artificial intelligence, automation and robotics