Rafay Launches Platform-as-a-Service Capabilities for GPU-based Workloads

Rafay Systems, a Platform-as-a-Service (PaaS) provider for modern infrastructure and accelerated computing, is extending the capabilities of its enterprise PaaS for modern infrastructure to support graphics processing unit- (GPU-) based workloads—making compute resources for AI instantly consumable by developers and data scientists.

Additionally, the company also launched a new AI Suite with standards-based pipelines for machine learning operations (MLOps) and large language model operations (LLMOps) to help enterprise platform teams quicken the development and deployment of AI applications for developers and data scientists, according to Rafay Systems.

The same issues that needed to be addressed for CPU-based workloads—environment standardization, self-service consumption of compute, secure use of multi-tenant environments, cost optimization, zero-trust connectivity enforcement and auditability—now must be addressed with GPU-based workloads. Aspects such as cost are even more critical to control in the new age of AI. 

In addition to applying its existing capabilities to GPU-based workloads, Rafay is extending its enterprise PaaS with features and capabilities that specifically support GPU workloads and infrastructure.

Rafay makes AI-focused compute resources instantly consumable by developers and data scientists, enabling customers to empower every developer and data scientist to accelerate the speed of AI-driven innovation—and do it within the guidelines and policies set forth by the enterprise, according to the company.

"I am immensely proud of Team Rafay for having extended our enterprise PaaS offering to now support GPU-based workloads in data centers and in all major public clouds,” said Haseeb Budhani, co-founder and CEO of Rafay Systems. “Beyond the multi-cluster matchmaking capabilities and other powerful PaaS features that deliver a self-service compute consumption experience for developers and data scientists, platform teams can also make users more productive with turnkey MLOps and LLMOps capabilities available on the Rafay platform. This announcement makes Rafay a must-have partner for enterprises, as well as GPU and sovereign cloud operators, looking to speed up modern application delivery.”

To address challenges associated with building and deploying AI-based applications, Rafay’s newly added support for GPU workloads helps enterprises and managed service providers power a new GPU-as-a-Service experience for internal developers and customers, respectively.

This provides developers and data scientists with:

  • Developer and Data Scientist Self-service: Easy to use, self-service experience to request for GPU-enabled workspaces
  • AI-optimized User Workspaces: Pre-configured workspaces for AI model development, training and servicing with necessary AI tools including Jupyter Notebooks and Virtual Studio Code (VSCode) internal developer environment (IDE) integrations
  • GPU Matchmaking: Similarly for CPUs, dynamically match the user workspaces with available GPUs or pools of GPUs based on criteria such as proximity, cost efficiency, GPU type and more to improve utilization
  • GPU Virtualization: Time slicing and multi-instance GPU sharing to virtualized GPUs across workloads and lower the costs of running GPU hardware with dashboards to visualize GPU usage

Rafay’s new AI Suite adds to Rafay’s existing portfolio of suites, which consists of the company's Standardization SuitePublic Cloud Suite, and Private Cloud Suite. New capabilities include:

  • Pre-configured LLMOps Playgrounds: Help developers experiment with generative AI (GenAI) by rapidly training, tuning, and testing GenAI apps with approved models, vector databases, inference servers, and more.
  • Turnkey MLOps Pipeline: Deliver an enhanced developer experience with an all-in-one MLOps pipeline, complete with GPU support, a company-wide model registry, and integrations with Jupyter Notebooks and VSCode IDEs.
  • Central Management of LLM Providers and Prompts: Built-in prompt compliance, cost controls on public LLM use such as OpenAI and Anthropic to ensure developers consistently comply with internal policies.
  • AI Data Source Integrations and Governance: Leverage pre-configured integrations with enterprise data sources such as Databricks and Snowflake while controlling usage for AI application development and deployments.

Rafay’s newly added support for GPU workloads also expands and enhances the solutions the company jointly brings to market with global partners such as NTT DATA. 

The new GPU-based capabilities in Rafay’s PaaS, along with the AI Suite are now generally available for customers. 

For more information about this news, visit