Instant Compute & Containers
Deploy VMs, containers, and Kubernetes clusters in under 30 seconds. Bare-metal performance with cloud flexibility, backed by our global high-speed network.
Choose Your Compute Model
From isolated VPS to fully managed orchestration, scale exactly how your application demands.
Standard VPS
Shared host CPU with guaranteed RAM & storage. Ideal for web apps, microservices, and development environments.
- 1-8 vCPU Cores
- 2-32 GB RAM
- Linux/Windows Images
- Private Networking
Dedicated CPU
Pinned physical cores for consistent performance. Database workloads, high-frequency trading, and batch processing.
- 2-32 Dedicated Cores
- Up to 256 GB RAM
- NUMA Optimization
- Low Latency Network
GPU Instances
A100 & L40S accelerators for AI/ML training, inference, and GPU rendering workloads.
- NVIDIA A100/L40S
- RDMA Networking
- Deep Learning Base
- Auto-Scaling
Container Instances
Run stateless containers without managing VMs. Pay only for vCPU and memory used per second.
- Docker/OCI Compatible
- Effortless Scaling
- Integrated Logging
- Health Checks
Container Groups
Co-located containers sharing storage and network. Perfect for sidecar patterns and complex services.
- Shared Volumes
- Host Network/Port
- Resource Quotas
- Init Containers
Serverless Containers
Zero-config scaling to zero. Event-driven compute with cold start optimization under 100ms.
- HTTP/gRPC Triggers
- Auto-Scale to 0
- Web Sockets
- Pay-per-Request
Managed K8s
Control plane fully managed by CloudNexus. Focus on workloads, not cluster maintenance.
- Auto-Patching
- Multi-AZ High Availability
- Integrates with GitOps
- RBAC & Policies
Service Mesh
Built-in Istio/Linkerd integration for traffic management, security, and observability.
- mTLS Encryption
- Canary Deployments
- Distributed Tracing
- Rate Limiting
Auto-Scaling Nodes
Cluster Autoscaler + Vertical Pod Autoscaler. Right-size your infrastructure automatically.
- Pod Priority Classes
- Spot Instance Integration
- GPU Auto-Scaling
- Cost Optimization
Developer-First Experience
Integrate with your existing toolchain. CLI, Terraform, API, and web UI.
$ cnx compute create --type k8s-cluster --region us-east-1
✓ Provisioning control plane...
✓ Attaching worker nodes (3x t3.large)...
✓ Configuring ingress & RBAC...
$ cnx kubectl get nodes
NAME STATUS ROLES AGE VERSION
worker-1 Ready 45s v1.28.3
worker-2 Ready 45s v1.28.3
worker-3 Ready 45s v1.28.3
$ _
Terraform & Pulumi
Infrastructure as Code with official, versioned providers. Destroy and recreate environments safely.
REST & GraphQL API
Full programmatic control. Every UI action is API-callable with consistent versioning.
CI/CD Integrations
Native plugins for GitHub Actions, GitLab CI, Jenkins, and ArgoCD. Zero-config deployments.
Docker Registry
Push images directly via `docker push`. 99.9% availability with geo-replication.
Technical Specifications
Enterprise-grade hardware with transparent resource allocation.
| Component | Specification | Notes |
|---|---|---|
| CPU Architecture | AMD EPYC™ 9004 / Intel® Xeon® Scalable | Latest gen, all-core turbo enabled |
| Memory | DDR5 ECC Registered (4800MHz) | Low latency, error-corrected |
| Storage | NVMe Gen4 SSD (Up to 7 GB/s) | Zero-copy I/O, thin-provisioned |
| Networking | 100 Gbps SmartNIC, SR-IOV | No noisy neighbors, dedicated VLANs |
| OS Support | Ubuntu, Debian, Alma, Rocky, RHEL, Windows | Custom ISO upload supported |
| Container Runtime | containerd, CRI-O, gVisor, Kata | Workload isolation options |
Compute Pricing
Pay only for what you use. No overage fees, no hidden costs.
Shared vCPU
Cost-effective for web & dev
- 1 vCPU / 1 GB RAM
- 25 GB NVMe SSD
- 1 TB Network Transfer
- Basic Monitoring
Dedicated vCPU
Performance-critical workloads
- 1 Dedicated Core / 4 GB RAM
- 80 GB NVMe SSD
- 4 TB Network Transfer
- Advanced Metrics & Logs
- Priority Support
GPU Accelerated
AI, ML & Rendering
- NVIDIA L4 / A100
- 32 GB vRAM
- 8 TB Network Transfer
- Deep Learning Base Image
- Spot Instances Available
Frequently Asked Questions
Everything you need to know about Compute & Containers.
Yes. Vertical scaling (RAM/CPU) for supported plans can be performed live with zero downtime. Horizontal scaling or OS changes may require a brief restart.
Each container gets a static private IP. Services can communicate via our integrated CNI plugin. External traffic is routed through configurable Load Balancers or Ingress controllers.
Daily automated snapshots are included at no extra cost. You can also create on-demand snapshots or configure continuous replication to another region for disaster recovery.
Yes. Spot instances reduce compute costs by up to 70%. They are ideal for fault-tolerant workloads, batch processing, and CI/CD runners. We provide 2-minute termination notices via webhook.
Ready to Deploy?
Get $200 in free compute credits. No credit card required. Spin up your first instance in 30 seconds.