Pure Storage & Cisco unveil unified platform for AI at scale
Pure Storage and Cisco have announced a new FlashStack Cisco Validated Design (CVD), developed in collaboration with NVIDIA, to provide enterprises with a unified production platform for deploying generative AI at scale.
The FlashStack CVD is an addition to Cisco's collection of AI PODs and forms a key module within the Cisco Secure AI Factory with NVIDIA. The product brings together compute, storage, networking, and software to help organisations transition from AI pilot projects to large-scale production deployments.
Addressing the challenges frequently faced by enterprises when operationalising AI, Maciej Kranz, General Manager, Enterprise, Pure Storage, said, "Too often, enterprises focus on GPUs and compute power, but without reliable data, the model never reaches its potential. Our collaboration with Cisco and NVIDIA removes those data barriers, giving customers the performance, simplicity, and efficiency they need to operationalise AI."
Organisations have been investing in pilot projects and proofs-of-concept, including testing large language models (LLMs), retrieval-augmented generation (RAG)-powered assistants, and other domain-specific AI tools. However, according to the involved companies, a common problem is that while models may be ready for production, the underlying data infrastructure may not be.
"With this new FlashStack CVD, we're not just validating hardware, we're orchestrating all the elements of RAG into an AI-ready infrastructure, removing complexity and reducing risk so customers can focus on turning data into insights that drive strategic outcomes," explains Jeremy Foster, SVP & GM, Cisco Compute.
Challenges in moving enterprise AI from pilot to production often stem from fragmented data architectures, limited visibility, and operational friction. The validated solution from Cisco and Pure Storage has been designed to overcome these obstacles by ensuring that both structured and unstructured data are available to AI workflows consistently and at scale.
Data integration
The solution incorporates Pure Storage's Enterprise Data Cloud architecture, which provides high-performance data access, concurrency, and improved energy efficiency through FlashBlade//STM. Portworx by Pure Storage facilitates persistent, portable, and protected data management across Kubernetes-based AI environments, giving enterprises the capability to support dynamic AI workloads.
Cisco and Pure Storage believe the integration allows AI teams to dedicate more of their efforts to innovation and less to infrastructure management. The combined platform aims to simplify the transition from AI experimentation to full-scale, reliable deployment across various use cases, including RAG, agentic AI, semantic search, video analytics, and code generation.
Enterprise-scale AI deployment
The validated solution features Pure Storage FlashBlade//STM for enterprise-level data management, Cisco UCS C845a servers with NVIDIA RTX PRO 6000 GPUs for GPU-accelerated compute, and NVIDIA AI Enterprise software for production-ready AI functionality. According to the companies, this configuration represents a comprehensive framework for converting raw data into the tokens that power generative AI systems in business environments.
By leveraging the experience of over 5,000 existing FlashStack customers, the newly launched FlashStack CVD integrates GPU-accelerated computing from Cisco, Nexus Switching, NVIDIA AI Enterprise software, and Pure Storage's data platform, eliminating the need for organisations to assemble disparate systems themselves.
Operational efficiency and control
AI workloads require seamless integration across compute, storage, and networking layers. The FlashStack CVD includes Cisco's Nexus 9000 series networking, delivering a low-latency, congestion-aware fabric tailored for high-performance AI tasks. Features such as telemetry, adaptive routing, and per-packet load balancing are designed to support efficient data flows between GPUs, storage, and compute nodes.
Using Cisco NX-OS and Nexus Dashboard, IT teams are provided with unified visibility and control, helping to maintain security standards while accommodating the specific requirements of AI-native workloads.
