Coredge Selects Lightbits to Power AI Cloud Services Infrastructure

Why It Matters

This shift shows how software‑defined storage can deliver cost‑effective, scalable AI infrastructure, accelerating cloud providers’ ability to serve regulated and high‑performance markets.

Key Takeaways

• Coredge adopts Lightbits SD storage for AI workloads
• Multi-petabyte deployment planned in India
• NVMe/TCP over Ethernet replaces legacy SAN
• Commodity hardware reduces total cost of ownership
• OpenShift/Kubernetes integration accelerates service rollout

Pulse Analysis

The surge in artificial‑intelligence workloads has placed unprecedented pressure on cloud providers to deliver storage that combines raw capacity with predictable latency. Traditional storage area networks (SANs) rely on specialized fabrics and proprietary appliances, inflating both capital and operational expenses. Lightbits Labs’ NVMe‑over‑TCP implementation leverages standard Ethernet, offering block‑level performance comparable to direct‑attached NVMe while retaining the flexibility of software‑defined architectures. This approach aligns with the broader industry shift toward disaggregated infrastructure, where compute and storage can scale independently. Enterprises can thus achieve petabyte‑scale storage pools without over‑provisioning network bandwidth.

Coredge, fresh from its acquisition by Sirius Digitech, is using this technology to build a multi‑petabyte AI platform in India. By deploying Lightbits’ software‑defined storage on commodity servers, Coredge eliminates the high TCO and rigidity associated with legacy SANs, while meeting the low‑latency, high‑throughput demands of AI training and inference. The solution integrates directly with the company’s OpenShift‑based Kubernetes environment, simplifying provisioning and enabling rapid service roll‑outs for regulated industries, telecom operators, and public‑sector clients. Early benchmarks suggest significant performance gains and faster time‑to‑market. Customers also benefit from built‑in data protection and fast snapshot recovery, aligning with strict SLAs.

The Coredge‑Lightbits partnership signals a broader market validation for NVMe‑over‑TCP as a cost‑effective alternative to traditional storage arrays. As AI models grow larger and latency becomes a competitive differentiator, cloud vendors that adopt software‑defined, Ethernet‑based storage can scale more nimbly and meet stringent compliance requirements. This deployment also highlights India’s emerging role as a hub for high‑performance cloud services, attracting investments that leverage local talent and infrastructure. Observers expect similar migrations across the sector, accelerating the move toward disaggregated, cloud‑native data fabrics. The move may also pressure legacy SAN vendors to innovate or risk losing market share.