FLEXPOD AI CISCO Oct 02, 2026
Speaking with theCUBE (SiliconANGLE) at NetApp INSIGHT 2026, NetApp chief partner and ecosystem officer Alvaro Celis and Cisco vice president of data center networking James Cole described FlexPod — the converged infrastructure franchise the two companies have sold jointly for 16 years — as a validated, pretested answer to AI’s integration complexity. Celis said the joint franchise has served 10,000 enterprise customers and generated $21 billion in ecosystem value. The pitch: enterprises want a trusted platform, not a pile of parts to integrate themselves.
“We have created an incredible alignment at the engineering level,” Celis said. “It’s not only Cisco and NetApp; it’s our shared channel that is actually making it real for customers so we can execute at scale.” Cole framed the change in how infrastructure is designed: “20 years ago, 10 years ago, you would’ve been designing your data structures and your applications around the infrastructure. Now we’ve turned that on its head… You’re really designing the infrastructure around the movement of data.”
The AI angle. The partners expanded FlexPod for AI deployments in June 2026, per NetApp’s investor newsroom. Celis said validated designs let the two companies align at the product, service, support and ecosystem level so customers get flexibility “with the certainty that it will work and it can work at scale.” He added that more joint FlexPod AI work is coming and additional companies are expected to join the partnership. Cole positioned the franchise’s two decades of iteration as its qualification for the next phase: “Now we’re taking that two decades of experience and bringing it to bear in AI, which is the most transformative infrastructure event we’ve had in many, many decades.”
Why it matters to ONTAP operators: a validated-design claim is a support-boundary claim, not a performance guarantee. FlexPod pairs ONTAP all-flash storage with Cisco compute and Nexus networking (and, in the AI references, NVIDIA GPUs), so treat “pretested” as a question to interrogate rather than a spec. Before standardising on a FlexPod for AI stack, pull the current validated design and confirm the exact ONTAP release, the NFS/pNFS and NVMe over TCP options the matrix actually qualifies, the tested failure and recovery behaviour, and which party owns the escalation path when the storage, fabric and GPU layers disagree. Our FlexPod reference covers the storage-layer building blocks, and the AI coverage index tracks the wider INSIGHT 2026 announcements. Note the evidence boundary: the source discloses no ONTAP version, benchmark, sizing or price, so none is asserted here.
Disclosure: theCUBE is a paid media partner for NetApp INSIGHT 2026; per SiliconANGLE, neither NetApp nor the event sponsors had editorial control over the content.
Read the original (SiliconANGLE / theCUBE) · NetApp–Cisco FlexPod for AI release (June 2026) · FlexPod reference · AI coverage index