Rafay Systems, a pioneer in delivering Platform-as-a-Service (PaaS) capabilities for self-service compute consumption, and Netris, a leader in networking automation, abstraction, and multi-tenancy for AI and cloud operators, announced a strategic partnership aimed at transforming GPU cloud infrastructure for enterprises racing to establish AI capabilities and applications.
Together, Rafay and Netris help accelerate consumption and monetization of GPU-based infrastructure by offering self-service workflows for model training, fine-tuning and inferencing use cases. This allows GPU cloud providers to transform raw GPU hardware into fully operational, enterprise-ready cloud platforms with comprehensive self-service capabilities in weeks — dramatically reducing the time-to-market and eliminating the complexity of building AI infrastructure.
GPU Cloud Market Challenges
Billions of dollars are being invested in high-performance GPU infrastructure — the market is expected to grow to $1.4 trillion by 2034. However, the current cloud offerings often lack essential self-service and multi-tenancy features to enable sharing of infrastructure across multiple tenants. These capabilities are fundamental to launching and operating a cloud service and are a top developer productivity priority for 44% of enterprises.
Building on lessons learned by leading cloud service providers, including Amazon Web Services, Microsoft Azure and Google Cloud Platform over the past decade, GPU cloud providers must prioritize self-service compute consumption to succeed. Without effective self-service workflows and appropriate infrastructure automation to support dynamic multi-tenancy, AI infrastructure adoption stalls, making meaningful return on investment difficult to achieve.
“Many first-generation GPU clouds still require human-driven backend processes to meet customer needs,” said Haseeb Budhani, CEO and co-founder of Rafay Systems. “Our partnership with Netris removes the traditional barriers of manual provisioning across the software and hardware layers of the AI infrastructure stack, allowing enterprises to instantly access and utilize GPU resources without the error-prone, human-in-the-loop processes that result in a less-than-ideal experience for consumers of AI infrastructure.”
Also Read: Next Pathway Enhances SHIFT with AI to Speed Data Migrations
“Big Three cloud providers invested a decade of top talent’s time to develop the proprietary software for infrastructure automation and multi-tenancy that enables their cloud offerings. Netris has built the same for everyone else covering the networking piece, while Rafay did the same for the compute piece,” said Alex Saroyan, CEO and co-founder of Netris. “Rafay and Netris together become a perfect turnkey solution for new and upcoming generations of GPU-based AI cloud providers.”
Rafay and Netris Bring Self-Service and Network Scalability to GPU Clouds
By combining Rafay’s enterprise-grade workflow management, virtualization and substrate management (also known as soft tenancy) with Netris’ advanced network lifecycle automation, abstraction and multi-tenancy (also known as hard tenancy), the partnership creates secure, isolated computing environments where multiple enterprises can leverage shared GPU cloud infrastructure without compromising performance, security or data privacy.
GPU cloud providers are empowered with the following benefits:
- Self-service consumption for developers and data scientists: Create customizable GPU/CPU packages and give developers and data scientists on-demand access to compute and AI applications with enterprise-grade security controls
- Multi-tenant infrastructure: Deploy secure, isolated environments with virtual cluster capabilities, network segmentation and policy enforcement for multiple enterprise customers
- Programmatic SKU management: Dynamically define and offer infrastructure packages tailored to market needs, from small training environments to large-scale inference deployments
- Integrated billing and chargeback: Access real-time usage data for seamless billing integration and financial visibility to maximize ROI on GPU investments
- End-to-end network automation: Configure networking layers automatically with built-in NVIDIA best practices to ensure hardware-level multi-tenancy, performance and scalability
- Guidance and support: Rafay and Netris each have more than six years of extensive experience supporting cloud service providers worldwide based on NVIDIA and other technologies
Source: Businesswire