Cisco pushes data fabric, production blueprints

Cisco used Splunk.conf25 in Boston on September 8, 2025 to introduce Cisco Data Fabric, a new architecture built on the Splunk platform for enterprise AI. (newsroom.cisco.com) The product is aimed at machine data — the logs, sensor readings, app events, network telemetry, and factory metrics companies already generate every day. Cisco said the system is designed to make that data easier to find, filter, tier, and use in AI models and agent workflows. (newsroom.cisco.com) Cisco said Data Fabric uses edge data management and federation so customers can analyze data where it already lives instead of moving everything into one place first. The company said that can support near real-time operational intelligence across multiple domains. (newsroom.cisco.com) (splunk.com) That pitch lines up with Cisco’s broader argument that enterprise AI is being blocked less by model availability than by data sprawl, infrastructure bottlenecks, and governance requirements. Jeetu Patel, Cisco’s president and chief product officer, framed those barriers as infrastructure, trust, and data gaps in Cisco and Splunk’s.conf25 messaging. (splunk.com) (crn.com) Cisco and Amazon Web Services are making the same case in current partnership materials that describe a shared “foundation” for AI across networking, cloud, security, and observability. Cisco’s AWS landing page says the goal is to help customers support AI and traditional workloads together without rebuilding from scratch. (cisco.com) A Cisco-AWS partnership brief says 74% of chief executives worry infrastructure gaps will leave them behind competitors, while 92% of companies plan to increase AI spending over the next three years. The same brief says only 1% of companies consider themselves mature in AI deployment. (cisco.com) Cisco is also packaging the production message into reference designs. Its AI PODs overview says customers can deploy pre-validated infrastructure in days rather than months, start at 32 graphics processing units, and scale to 128 or more GPUs per cluster. (cisco.com) In a separate Cisco-AWS manufacturing brief, Cisco says 56% of manufacturers have hit data-related problems in AI projects and that downtime costs the sector $255 million a year. The document argues AI systems in factories need reliability, uptime, and governance from the start because they run inside live production environments. (cisco.com) Cisco’s own research and sponsored materials show the pressure behind that push. A Cisco-backed MIT Technology Review Insights report says 98% of companies feel more urgency around AI than a year ago, 85% think they have less than 18 months to set an AI strategy, and only 13% say they are ready to use AI to its full potential. (cisco.com) Taken together, the current Cisco story is less about unveiling one more model and more about building the plumbing around enterprise AI: data access through Splunk, validated infrastructure through AI PODs, and joint cloud-and-network controls with AWS. (newsroom.cisco.com) (cisco.com)