NEW YORK, February 13, 2024 — VAST data today announced a partnership with Run:ai, a leader in compute orchestration for AI workloads. This collaboration marks a monumental step in redefining large-scale AI operations, delivering a full-stack solution that includes compute, storage, and data management. Together, VAST and Run:ai address a critical need for companies embarking on large-scale AI initiatives.
Run:ai streamlines accelerated NVIDIA AI infrastructure across private, public, and hybrid clouds while increasing AI project efficiency through dynamic workload scheduling and innovative GPU partitioning to improve GPU allocation and Improve utilization. The platform addresses a variety of needs, from supporting interactive environments for data scientists to improving training at scale and reliable, scalable inference. Run:ai’s open architecture ensures a future-proof collaborative platform that integrates with a broad ecosystem of industry leaders. For the data layer, the VAST data platform combines storage, database, and containerized compute engine services into a single platform built from the ground up to power AI and GPU acceleration tools in modern data centers and clouds. Integrate into a scalable software platform.
Renen Hallak, CEO and co-founder of VAST Data, said: “Our partnership with Run:ai goes beyond traditional disparate AI solutions and brings together all the components needed for an efficient AI pipeline. We provide a blueprint for delivering targeted, effective, and innovative AI operations at scale.”
VAST Data and Run:ai work together to provide organizations with the following capabilities:
- Full stack visibility for resource and data management: The synergy between the VAST data platform and Run:ai creates a comprehensive AI solution that provides enterprises with full-stack visibility including compute, networking, storage, and workload management across AI operations.
- Cloud service provider ready infrastructure: CSP is critical to bringing GPU availability to companies integrating AI into their business processes. VAST and Run:ai provide CSPs with a blueprint to efficiently deploy and manage AI cloud environments. The VAST data platform, in conjunction with Run:ai, enables CSPs to leverage secure enterprise-grade AI across a single, shared infrastructure using a zero-trust approach for compute and better data isolation and utilization. We provide a tested and validated framework for providing environments.
- Optimized end-to-end AI pipeline: From multiprotocol ingestion to data processing to model training and inference, organizations can use NVIDIA RAPIDS accelerators for Apache Spark as well as others available with the NVIDIA AI Enterprise Software Platform for production development and deployment. Accelerate data preparation using AI frameworks and libraries. -Grade AI application. Enable high-performance data preprocessing using the VAST database.
- Simple AI deployment and infrastructure management: Combined with simplified multi-GPU distributed training, Run:ai offers fair-share scheduling that allows users to easily and automatically share clusters of GPUs without memory overflow or processing conflicts. To do. Additionally, VAST DataSpace facilitates data access across geographies and multi-cloud environments while providing the encryption, access-based controls, and data security that customers require.
The VAST Data Platform is VAST’s breakthrough approach to data management and accelerated computing. For enterprises and CSPs, the platform provides the comprehensive software infrastructure needed to capture, catalog, refine, enrich, store, and protect unstructured data using AI and deep learning real-time analytics. It works. Through the Run:ai open architecture, Run:ai seamlessly integrates with NVIDIA AI Enterprise and NVIDIA Accelerated Computing to help customers accelerate development, scale AI infrastructure, reduce compute costs, and increase management resources to effectively coordinate and manage them.
By deeply integrating NVIDIA’s market-leading AI computing products with the dynamic AI workload orchestration of the Run:ai platform and VAST’s industry-disrupting AI data platform, organizations can make optimal use of their resources. , gain greater control and visibility across both the compute and data tiers. .
VAST + Run:ai blueprints, solution briefs, and demos will initially be available at: NVIDIA GTC 2024. To learn more about VAST and Run:ai on NVIDIA GTC, access VAST data at: Booth number 1424.
Source: VAST data
