Compute
On-demand virtual machines
Provision and manage instances
Flexibly allocate vCPUs and memory, and combine with an OS image. Import your own images or build them with tools like Packer.
Seamless SSH key integration
Flexible image management
Self-service instance networking
Automate resource provisioning
Unified interface: everything is available via API, CLI and web console. Native Terraform integration.
Integrated metrics and telemetry
Monitor CPU, memory, disk & network metrics from a single control plane.
Project-level resource limits
Per-project and per-tenant quotas to prevent resource overconsumption.
2nd Gen Oxide Compute
AMD EPYC 9005 Series processors bring Zen 5 or Zen 5c cores, DDR5 memory, and dramatically higher throughput to every sled in the rack. A generational leap in performance and efficiency for the AI, Cloud, and Enterprise Computing workloads that matter most.
Leadership AVX-512 Performance
AVX-512 support for workloads that depend on wide vector operations — AI inference, fluid dynamics, financial modeling, signal processing — without specialized hardware.
- AI inference and vector-heavy computation
- Lower latency for sensitive workloads
- No specialized hardware required
More useful work per system
Oxide's integrated design eliminates the overhead of legacy rack architecture so more of your power envelope and floor space go toward running real workloads.
- Massively parallel compute capacity per sled
- High workload density per system
- Per-node capability that optimizes fleet size
Hardware and software co-design
Rack-scale platform, fully integrated, software-complete and API-first for true cloud-native operations.
- One unified control plane
- Metrics API for VMs and hardware
- Debug with full-stack observability