Sure! Here’s the translation:
The integration of WEKA’s data platform with the Nebius cloud offers next-generation GPUaaS to accelerate the training and inference of artificial intelligence models.
WEKA, a company specializing in native data platforms for AI, and Nebius, a leading provider of cloud infrastructure for AI, have announced a strategic alliance to offer a high-performance GPU-as-a-Service (GPUaaS) solution targeted at organizations that need to scale their computing and storage resources on demand, with microsecond latencies for advanced AI projects.
The collaboration integrates WEKA’s high-performance storage software with Nebius’s AI cloud platform, allowing businesses, startups, and research centers to efficiently scale from petabytes to exabytes of data, optimizing the training and inference of AI models and facilitating the development of innovative projects in areas such as generative AI and autonomous agents.
An optimized neocloud for advanced AI
Nebius AI Cloud provides a cutting-edge cloud environment, cost-optimized and specifically designed for large-scale AI workloads. In this context, Nebius has selected WEKA’s storage solution to enhance its premium platform, highlighting its ability to scale and manage large volumes of data with exceptional performance, both in read and write operations, along with advanced metadata management and multitenancy.
One of the first success stories has been the deployment of the solution in a leading research center, which utilizes Nebius’s infrastructure for large-scale experiments and model development, featuring a cluster of several thousand GPUs and 2 petabytes of integrated WEKA storage. This architecture facilitates user and directory quota management, as well as the availability of a fully managed, robust, and flexible platform to meet the demanding requirements of AI research.
"Maximum performance and operational simplicity"
Danila Shtan, CTO of Nebius, noted that WEKA’s solution has exceeded all expectations: "Not only does it deliver outstanding performance in throughput, IOPS, and low latency at scale, but it also efficiently manages mixed workloads of reads and writes, along with very advanced metadata and multitenancy handling."
Meanwhile, Liran Zvibel, co-founder and CEO of WEKA, emphasizes: "We take pride in collaborating with Nebius to provide high-performance cloud solutions that maximize AI innovation while minimizing infrastructure complexity. Together, we are redefining what’s possible when high-performance storage meets an infrastructure truly designed for AI."
Cloud for the new era of artificial intelligence
Based in Amsterdam and present in Europe, North America, and Israel, Nebius has built its AI Cloud platform from the ground up for intensive AI workloads, combining proprietary cloud architecture, in-house designed hardware, and managed services for developers.
On its part, WEKA has reinvented the enterprise data stack for the era of agent-based AI, enabling data portability across on-premises, cloud, and edge environments, eliminating silos, and transforming them into dynamic pipelines to maximize GPU utilization and accelerate the time-to-market for new models and solutions.
The WEKA-Nebius alliance marks a milestone in the provision of custom cloud infrastructure for artificial intelligence, combining scalability, extreme performance, and ease of management to enable the next generation of AI applications in both business and advanced research.