01 / Bare-metal
Dedicated Servers
Custom CPU, RAM, GPU & bandwidth — single-tenant power, deployed instantly with full root access.
Configure →Dedicated · Bare-metal · GPU
Out-of-band management on every dedicated server — take the console, cycle power, watch every sensor. Total command of your bare metal, from anywhere.
New York · Miami · San Francisco · Amsterdam · Bucharest
Trusted at scale
Uptime SLA
Global regions
Human support
Years online
Five regions, one network
Built for your workload
01 / Bare-metal
Custom CPU, RAM, GPU & bandwidth — single-tenant power, deployed instantly with full root access.
Configure →02 / Compute
NVIDIA & AMD acceleration for AI training, LLM inference, transcoding and computer vision.
See GPUs →03 / Network
1–100 Gbps unshared ports with flat, predictable billing — built for streaming, CDN and data-heavy pipelines.
Get unmetered →New service
Run OpenClaw on enterprise-grade bare metal — pre-configured with your OS and OpenClaw installed, ready to deploy in minutes. From $34.40/mo.
| Plan type | Best for | Network | Storage | Starts at |
|---|---|---|---|---|
| Dedicated / Bare Metal Servers | Web apps, databases, virtualization, containers, private clouds, custom OS | 1-100 Gbps | Enterprise SSD / NVMe / HDD | $5/mo |
| GPU Dedicated Servers | AI/ML, transcoding, computer vision | 1-100 Gbps | Enterprise SSD / NVMe / HDD | $59/mo |
| Unmetered Servers | Streaming, CDN, data transfer at scale | 1-100 Gbps unshared | Enterprise SSD / NVMe / HDD | $529/mo (10 Gbps unmetered) |
Build infrastructure exactly to spec with bare-metal servers that behave like owned hardware, not segmented slices of a multi-tenant platform. Allocate resources without artificial limits, attach GPUs, or expand bandwidth as traffic grows. Use Primcast’s high-capacity backbone to sustain long-term throughput without throttling, shaping, or unpredictable billing events.
Shift production workloads using controlled migration handled by infrastructure specialists. Deploy under a 99.9% uptime SLA that prioritizes live environments and bypasses traditional ticket queues for immediate engineering support. Start with instant inventory, or on a custom deployment: specific CPUs/GPUs, private racks, cross-connects and more.