Backblaze is rolling out a new cloud object storage product aimed directly at neocloud platforms, the companies building AI-focused cloud environments for high-performance computing.
The move comes as AI workloads increasingly strain storage infrastructure. Training large models, running inference, and processing media-heavy data all require systems that can keep pace with powerful GPUs. For many neocloud providers, building that storage layer internally requires committing significant capital and engineering talent that might otherwise be used to expand compute capacity. Instead of starting from scratch, they can integrate B2 Neo behind the scenes and present it to customers as part of their own platform.
Backblaze says the offering is built on nearly 20 years of operating cloud storage at scale. The company manages more than five exabytes of data and supports a throughput of up to 1 terabit per second. B2 Neo was developed alongside several neocloud platforms already running production workloads on its infrastructure. Those early collaborations include the company’s largest total contract value commitment to date, as well as active discussions with other emerging providers seeking to strengthen their storage capabilities.
Operational Integration and Platform Control
The structure of the offering is intended to give neoclouds operational control while outsourcing the complexity of storage infrastructure. Customers of those platforms will see storage presented as a native service, complete with branded endpoints and pricing determined by the partner. Account provisioning, permission management, and billing can be handled through neocloud’s existing tools, eliminating the need for a separate administrative console.
Market forecasts suggest significant growth headroom. The neocloud segment is projected to expand from $35.22 billion in 2026 to $236.53 billion by 2031, representing a compound annual growth rate of 46.37%. As these platforms race to add GPU capacity, storage has emerged as a critical bottleneck. Engineering teams often face a trade-off: allocate resources to building scalable object storage or concentrate on expanding compute infrastructure that differentiates their offerings.
For companies running AI training pipelines, inference jobs, or large-scale media processing, storage is not just a background utility; it is central to keeping everything moving. Massive datasets, model checkpoints, and output files must be stored and accessed quickly and reliably. When object storage is not built directly into a platform, teams often end up shuttling huge volumes of data back and forth between environments. That extra movement can introduce latency, leave expensive GPUs idle, and drive up overall costs.
Strategic Focus on Compute Scaling
Backblaze positions the new offering as a way to ease that pressure. Neocloud providers are racing to add GPU capacity, and building high-performance object storage in-house can pull focus and resources away from that mission. By offering a ready-made storage layer that can be deployed in weeks rather than years, the idea is to let these platforms concentrate on scaling compute while still delivering integrated storage to customers.
Availability
Backblaze B2 Neo is now available to qualified neocloud platforms.
Amazon