Whole-machine performance without the hypervisor tax. Out-of-band management, hardware RAID, and custom OS images on every node — billed monthly, racked in minutes, not weeks.
Provisioned in 30 minutesIPMI + BMC out-of-bandHardware RAID + custom OS
01Starter
Intel Xeon E-2388G
Entry-tier production
8 cores / 16 threadsBoost 5.1 GHz
32 GB DDR4 ECC
2× 512 GB NVMeHardware RAID 1
1 Gbps uplink10 TB egress / mo
Web and app tiers, single-instance Postgres or MySQL, CI runners, and lightweight Kubernetes nodes.
From
$99/mo
02Most popular
AMD Ryzen 9 7950X
High-clock production
16 cores / 32 threadsBoost 5.7 GHz
64 GB DDR5
2× 1 TB NVMe Gen4Software RAID
1 Gbps uplink20 TB egress / mo
Real-time SaaS backends, multiplayer game servers, parallel build farms, and video transcoding pipelines.
From
$179/mo
03Enterprise
AMD EPYC 9354P
Scale-out enterprise
32 cores / 64 threadsBoost 3.8 GHz
256 GB DDR5 ECC
2× 3.84 TB NVMeHardware RAID 10
10 Gbps uplink50 TB egress / mo
Multi-tenant SaaS platforms, OLAP and analytics clusters, in-house ML inference, and large multiplayer fleets.
Everything you need to know about our platform and services.
We offer shared vCPU, dedicated vCPU, and bare-metal server options. Shared instances are ideal for dev/staging, dedicated vCPU for production workloads, and bare-metal for maximum performance with no virtualization overhead.
All major Linux distributions including Ubuntu, Debian, CentOS, Rocky Linux, AlmaLinux, and Fedora. Windows Server images are also available. You can also upload and boot custom ISO images.
Most instances are provisioned within 30-60 seconds. Select your region, OS, and plan — your server will be ready with a public IP and full root/SSH access.
Yes. You can vertically scale (upgrade vCPU, RAM, and storage) with minimal downtime. Horizontal scaling is also supported through our load balancer and auto-scaling groups.
All instances use enterprise-grade NVMe SSDs with up to 7 GB/s read speeds. Block storage volumes can be attached for additional capacity and are replicated across three nodes for durability.
Each plan includes a generous bandwidth allowance (4 TB to 32 TB depending on tier). Inbound traffic is always free. Overages are billed at competitive per-GB rates with no surprises.
Yes. All compute instances include always-on L3/L4 DDoS mitigation at no extra cost. L7 protection and advanced WAF rules are available as add-ons.
Use cases
Compute that fits the workload.
Four workloads our customers ship every day — same NVMe instances, 12-region footprint, 24/7 support.
01
Web & SaaS
Production web apps and APIs
Host high-traffic sites, REST/gRPC APIs, and SaaS backends on dedicated cores with sub-20 ms regional latency, managed TLS, and instant blue-green deploys.
02
Data tier
Self-managed database servers
Run PostgreSQL, MySQL, Redis, or MongoDB on NVMe-backed instances with snapshot scheduling, point-in-time recovery, and replicated block volumes.
03
CI/CD
Build farms and CI runners
Parallel build pipelines and self-hosted runners on dedicated CPUs — persistent cache volumes, ephemeral worker pools, and BYO container registry.
04
Real-time
Game, media, and streaming servers
Low-latency multiplayer backends and live media origins across 12 regions, with always-on L3/L4 DDoS mitigation and 25 Gbps networking.