Scality and WEKA announced a new jointly validated solution that combines WEKA’s NeuralMesh high-performance storage with Scality RING’s cost-efficient object tier. The solution offers enterprises the best of both worlds: flash-speed performance for active AI and high-performance computing (HPC) data, combined with scalable, durable capacity for large datasets, while reducing cost and complexity without compromising speed.
At the heart of the new solution is Scality’s proven object connector for NeuralMesh, a lightweight, object storage interface validated by WEKA at scale. Compared to conventional S3 interfaces, Scality RING achieved up to 10x faster performance on similar hardware in Scality testing, with up to 20% lower infrastructure costs. This enables organizations to extend their NeuralMesh-powered AI and HPC data pipelines more economically while avoiding the management overhead associated with traditional object stores.
“WEKA’s NeuralMesh storage system delivers the high-performance software foundation modern AI pipelines require to run optimally. Enterprises that leverage the Scality RING lightweight connector together with NeuralMesh can achieve additional economic benefits, leveraging a cost-efficient object tier,” said Nilesh Patel, chief strategy officer at WEKA. “This enables our mutual customers to achieve their AI project outcomes faster and more efficiently.”
The joint solution architecture keeps new and active data on WEKA’s NeuralMesh flash tier, while seamlessly tiering data to Scality’s exabyte-scale, resilient object storage. This ensures optimal performance and long-term cost control, without forcing customers into expensive, all-flash deployments.
Customer benefits include:
- Performance when and where it matters: NeuralMesh by WEKA maximizes GPU utilization, accelerates time to first token, and powers AI pipelines at flash speed.
- Economics at scale: Scality RING scales to exabytes with up to 14 nines durability, with the object connector for NeuralMesh providing an efficient object tier with up to 20% lower costs compared to traditional object integration.
- Simplified tiering: Proven interoperability validated by WEKA – no engineering changes required.
- Speed and reliability vs. Ceph: Purpose-built for NeuralMesh by WEKA, the new connector integration provides a faster, more manageable, and reliable alternative to community-driven object stores, with enterprise-grade support.
“Our partnership with WEKA gives enterprises a smarter way to extend the economics of their AI pipelines,” said Erwan Girard, chief product officer at Scality. “WEKA drives the performance; Scality provides the scale. Together, we help customers reduce infrastructure costs and management complexity while keeping their AI and HPC environments running at peak efficiency.”
Availability: The Scality RING with NeuralMesh Connector for WEKA solution is available now.
