Dedicated hardware tailored to your application
Run workloads on dedicated capacity
Use customer-dedicated servers as Kubernetes nodes or standalone systems beside the cluster.
Keep hybrid systems fast and private
Connect Kubernetes, databases, and dedicated servers over high-speed PrivateLAN instead of public networks.
Run workloads on dedicated capacity
Use customer-dedicated servers as Kubernetes nodes or standalone systems beside the cluster.
Keep hybrid systems fast and private
Connect Kubernetes, databases, and dedicated servers over high-speed PrivateLAN instead of public networks.
Node technical specs
Three master nodes for high availability, with regular backups.
| CPU cores | RAM | Network | System SSD | Data SSD | |
K-CP-310 - Master node | 8 | 32 GB | 10G | 240 GB | 480 GB |
Multifunction nodes for common application workloads and legacy-compatible use cases.
| CPU cores | RAM | Network | System SSD | User SSD | |
K-WN-110-G - Versatile computing node | 20 | 64 GB | 20G/25G | 240 GB | 480 GB |
K-WN-120-G - Large computing node | 32 | 128 GB | 20G/25G | 240 GB | 960 GB |
For distributed single-tenant file and object storage, event streaming, and messaging.
| CPU cores | RAM | Network | System SSD | Storage SSD | |
K-WN-180-S - Compact storage node | 20 | 64 GB | 25G | 240 GB | 2 x 480 GB |
K-WN-180-S.10 - Medium storage node | 20 | 64 GB | 25G | 240 GB | 2 x 960 GB |
K-WN-190-S - Extra-large storage node | 20 | 128 GB | 25G | 240 GB | 2 x 3840 GB |
For CPU-intensive workloads, analytics, scientific computing, and high single-threaded performance.
| CPU cores | RAM | Network | System SSD | User SSD | |
K-WN-20-C - Medium computing node | 20 | 32 GB | 20G/25G | 240 GB | - |
K-WN-30-C - Large computing node | 32 | 64 GB | 20G/25G | 240 GB | - |
K-WN-40-C - High-performance computing node | 12 | 48 GB | 20G/25G | 240 GB | 480 GB |
Connect Nvidia / AMD GPUs in your cluster.
Connect Nvidia / AMD GPUs in your cluster
We configure GPU passthrough and driver management on dedicated nodes for inference, training, and AI pipelines.
Contact usFor workloads outside Kubernetes that still need private connectivity to the cluster.
Fully customisable bare-metal servers
KVM-over-IP for remote management, same private network as your Kubernetes cluster. Build the exact configuration your workload needs.
See dedicated serversFAQ
Your infrastructure backlog, handled.
From first conversation to running cluster, typically five working days. Fixed monthly price, dedicated hardware, engineers you can reach.