Rafay-powered Virtual Machines-as-a-Service (VMaaS)

Rafay-powered Virtual Machines as a Service (VMaaS) delivers GPU- and CPU-based virtual machines as secure, scalable, consumption-based services, built on mature KVM virtualization. Rafay turns existing compute into fully managed, tenant-ready VM SKUs with governance, visibility, and automated lifecycle control, with NUMA topology-aware passthrough for near bare-metal performance and hardware-level tenant isolation.

• Near Bare-Metal Performance: NUMA topology-aware GPU passthrough on KVM, validated against the NVIDIA reference architecture.
• Hardware-Level Isolation: NVIDIA DPU-based hard tenancy with SELinux-enforced boundaries and per-tenant network and storage isolation.
• On-Demand with Metering: Self-provision 1 to 8 GPUs per VM, with per-second metering and billing.

Simplify Virtual Machine Provisioning and Governance

One workflow places, provisions, and configures each VM across environments

Smart Scheduling

Intelligent matchmaking places each VM by NUMA topology and GPU layout, then provisions it in one workflow

Integrated Storage Options

Turnkey connections to Ceph, DDN, Weka, Vast, Dell PowerStore, and other platforms

Hardware-Level Multi-Tenancy

NVIDIA DPU-based hard tenancy, with TEE-secured confidential VMs for sensitive workloads

Network Automation

Per-tenant VRF isolation, public and private IPAM, and automated firewall and NAT rules, with dynamic VRFs via Netris.

Convert Infrastructure into Compliant, High-Value Services

Convert idle compute resources into cloud-grade virtual machine services in weeks.

NUMA-aware passthrough delivers full GPU performance with the flexibility of VMs.

Deliver compliant, in-region VM services with VPC isolation and built-in governance.

Flexible SKUs and vGPU sharing drive fleet density, with per-second metering for precise chargeback.

Featured Resources

• Operationalizing AI Fabrics with Aviz ONES, NVIDIA Spectrum-X, and Rafay
  Discover the new AI operations model.

• The Definitive GPU PaaS Reference Architecture
  Understand what it takes to deliver the right GPU infrastructure.

• Unlock Your AI Potential with Cisco and Rafay: Transform AI PODs into a Self-Service GPU Cloud
  Cisco provides AI-optimized infrastructure.

• The CIO’s guide to scalable, compliant, and developer-ready AI deployment
  Orchestrating the future of AI.

• Rafay Named Outperformer in 2025 GigaOm Radar Report for Managed Kubernetes
  Ranked as an “Outperformer” for its solution.

• Building AI Value within Borders
  Facilitates efficient, self-service infrastructure.

• GPU cloud evaluation report
  Evaluating the Rafay Platform.

• How Enterprise Platform Teams Can Accelerate AI/ML Initiatives
  Key challenges and best practices.