Entry 1 — Ryzen 5 3600
Six cores and 64 GB of memory for websites, small databases and test environments.
- Cores
- 6
- Threads
- 12
- DDR4
- 64 GB
- Traffic
- 20 TB/mo
- Storage
- 2 × 512 GB NVMe (RAID 1)
- Uplink
- 1 Gbit/s
- Location
- Dubai, United Arab Emirates
Pre-configured physical servers — one tenant, one chassis, no hypervisor in between. Every spec below is the hardware we rack, and the stock figure is the number we have. Want to choose each component instead? Advanced Setup →
Single-socket machines for websites, small databases and test environments.
Six cores and 64 GB of memory for websites, small databases and test environments.
Eight cores with DDR5 ECC memory in a compact single-socket server.
A six-core Xeon with ECC memory for one business application.
AMD EPYC with DDR5 ECC memory in a compact single-socket server.
Eight Xeon cores and 64 GB of ECC memory, with mirrored drives.
More cores and memory for busy applications, databases and build servers.
Eight cores at up to 5.3 GHz for applications that need fast single cores.
24 cores (8 performance + 16 efficient) and 64 GB of DDR5 ECC memory.
Twelve EPYC cores with DDR5 ECC memory for busy applications and databases.
16 cores with AMD 3D V-Cache and 128 GB of DDR5 ECC memory.
16 EPYC cores with 3D V-Cache, 128 GB of ECC memory and a 10 Gbit/s uplink.
48 cores and 256 GB of DDR5 ECC memory in one socket, with a 10 Gbit/s uplink.
Two 32-core Xeons and 512 GB of ECC memory for large virtualisation and database hosts.
Large RAID arrays for backups, archives, media and file storage.
32 TB usable in RAID 10, with 64 GB of ECC memory.
132 TB usable on eight drives in RAID 6, with 128 GB of ECC memory.
30.7 TB usable flash storage in RAID 10, with a 10 Gbit/s uplink.
264 TB usable on fourteen drives in RAID 6, with 32 cores and a 10 Gbit/s uplink.
NVIDIA GPUs for AI inference, rendering and video work.
One NVIDIA RTX 4000 SFF Ada (20 GB) for inference, rendering and video encoding.
One NVIDIA L4 (24 GB) with an EPYC host and 128 GB of ECC memory for AI inference.
One NVIDIA RTX 6000 Ada (48 GB) on a 24-core Xeon host for training and 3D work.
Two NVIDIA L40S (48 GB each), two 32-core EPYCs and 512 GB of ECC memory.
Single-socket machines for websites, small databases and test environments.
A single-socket workhorse for one busy application.
High clocks per core, for workloads that stay single-threaded.
More cores and memory for busy applications, databases and build servers.
Eight cores and a fast mirrored NVMe pair.
Two sockets, 24 cores, and room on the disk shelf.
32 cores across two sockets, with ECC memory to match.
48 cores, DDR4 ECC and an all-NVMe array.
Physical machines are prepared by hand. Ordering reserves one of the chassis above and raises an invoice; once it is paid our team racks the server and sends you the handover details. Nothing is provisioned automatically, and we will tell you where it is up to.
Prices can change at any time — we work to keep them among the most affordable. A service you've already bought keeps the price you ordered it at.
An automated assistant that answers from our help pages, FAQs and plans. It can make mistakes. Contact support for anything it cannot answer.
Browse our guides or reach our support team. We typically reply by email during business hours.