AI Is Reshaping Your Entire Stack, Not Just Applications

Apr 2, 2026

Mano Bhattacharyya, Nutanix CTO and enterprise infrastructure expert, explains why AI reshapes compute, data, security, and economics across the stack. He covers GPUs, ARM, DPUs, bare-metal VMs, dirty data and knowledge graphs, agent security and gateways, and why cross-functional AI committees and use-case driven prototypes beat siloed experimentation.

Ask episode

AI Snips

Chapters

Transcript

Episode notes

INSIGHT

AI Forces An End-to-End Infrastructure Shift

AI changes the entire infrastructure stack from hardware to application, not just the application layer.
Mano Bhattacharyya highlights a shift from x86-centric servers to GPUs, ARM-based CPU+GPU integrations, and new hypervisors tuned for AI workloads.

INSIGHT

DPUs And GPU Direct Rewrite I/O Patterns

Networking and storage are evolving with DPUs and GPU-direct I/O that offload work from CPUs.
Mano explains NICs are no longer ‘dumb’ and that hypervisors must optimize I/O and GPU access for bare-metal VM performance.

INSIGHT

Dirty Data Blocks AI Value

Data quality and management are the lifeblood constraint for enterprise AI deployments.
Mano calls out unclean, unsecured, and unjoined enterprise data and rising use of knowledge graphs to relate structured and contractual unstructured data.

Get the Snipd Podcast app to discover more snips from this episode

Get the app

AI is forcing enterprises to rethink everything from hardware to governance, and most organizations are attacking it in silos.

Mano Bhattacharyya (CTO, Nutanix) breaks down why AI isn't just an application layer problem, but an end-to-end transformation that spans compute (GPUs, ARM, DPUs), data management (unclean enterprise data, knowledge graphs), security (agent gateways, MCP server risks), and economics (token costs vs. usage explosion). He explains why CIOs need cross-functional AI committees, not isolated strategies, and why use case driven AI beats exploratory projects that burn budgets in months.

The solution is to form AI committees where CIO, CTO, and CDO work together, not in silos. Focus on use case driven AI by learning from peers, rather than exploratory AI that becomes a budget trap. Start with small prototypes with dedicated use cases, not a free-for-all where every team tries something.

Chapters:

0:00 AI as an End-to-End Infrastructure Challenge

2:16 Networking, Storage, and Bare Metal VM Performance

4:43 Agent Security, Gateways, and Enterprise Governance

6:33 Use Case Driven AI and Learning from Peers

This episode of IT Visionaries is brought to you by Meter - the company building better networks. Businesses today are frustrated with outdated providers, rigid pricing, and fragmented tools. Meter changes that with a single integrated solution that covers everything wired, wireless, and even cellular networking. They design the hardware, write the firmware, build the software, and manage it all so your team doesn't have to.

That means you get fast, secure, and scalable connectivity without the complexity of juggling multiple providers. Thanks to meter for sponsoring. Go to meter.com/itv to book a demo.

---

IT Visionaries is made by the team at Mission.org. Learn more about our media studio and network of podcasts at mission.org.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.