Four locations,
one way of working.

Pick whichever is closest to your users. Whichever you pick, the port has no traffic counter on it and the machine is yours alone.

IADPHX
// united statesIAD Manassas, VirginiaPHX Phoenix, Arizona
AMSMEP
// netherlandsAMS AmsterdamMEP Meppel

city, not building · the two maps are not to the same scale

∞
traffic · no cap, no overage
100Gbps
largest port per host
4
locations · AMS · MEP · IAD · PHX
v4+v6
2 ipv4 included · ipv6 native

Four named buildings, our own hardware.

The machines stand in buildings of Tier III and Tier IV design, run by the operators below, on hardware we own and keep spares for. Nothing here is a promise about uptime: we sign no percentage. It describes what the building and the box are built from.

buildingcitycertifications
Equinix AM3AmsterdamISO 9001 · ISO 14001 · ISO 22301 · ISO 27001 · ISO 45001 · ISO 50001 · SOC 1 Type II · SOC 2 Type II · PCI DSS · TIA-942
Kolo NL2, formerly ServeriusMeppelISO 9001 · ISO 27001
Anexia ANX85Manassas, VirginiaISO 9001 · ISO 14001 · ISO 27001 · ISO 27701 (Anexia, company-wide)
phoenixNAPPhoenix, ArizonaSOC 1 · SOC 2 · PCI DSS · HIPAA-ready · AZRamp

Certifications as each operator states them, checked on 2 October 2026.

facilities
data centres of Tier III and Tier IV design
power
two power supplies in every server, redundant power to the rack, diesel generators on site as a rule
cooling
redundant cooling, as the Tier III and IV classes require
access
controlled physical access, as the Tier III and IV classes require
hardware
Supermicro, Dell, Gigabyte; all of the hardware is ours, with a stock of spare parts
memory
ECC memory in every server
replacement
a failed part is replaced from that stock at our cost; a machine that is down is an emergency, handled as fast as possible, often within hours
on site
technicians on site in all four data centres, 24/7, weekends and holidays included
remote hands
failed hardware is ours to fix, at our cost; other physical work you ask for is priced per job: rates differ between data centres and change over time, so we quote before we start
before handover
every server is fully tested before handover: memory, disks and ports

Any port, in any of the four.

In each of these buildings we can hand you a server on a 1, 10, 25, 40 or 100 Gbps port. Every port is unmetered: no allowance, no overage, no 95th percentile. Stock servers go up to 10 Gbps; anything above that is a custom build. Which stock builds are available, and at what price, depends on the location.

Don’t take our word for it.

Run these on the machine once it is yours. No results here: the ones from your server are the ones that count.

verify.logexample commands
# run on your own server, once it is yours
$ iperf3 -c <your-host> -P 16 -t 86400
#  a day at line rate; compare with the port you ordered
$ mtr -rwc 200 <your-host>
#  path and loss from where your users are
$ ipmitool -I lanplus -H <ipmi-address> -U <user> chassis status
#  out-of-band access answers to you, not to us
$ nproc; free -g; lsblk
#  cores, memory and disks match the order

We do not sell DDoS protection.

There is no scrubbing product here, no advertised mitigation capacity and no promise about what we can absorb, because there is no service behind those words on our side.

what is there
filtering at the network edgemonitoring that pages a humana saturated port is never treated as suspicious
what is not
no scrubbing centre or mitigation tierno advertised absorption capacityno guarantee about a given attack

If your workload attracts attacks and you need guaranteed mitigation, say so before you order. It is a fair reason for us not to be the right host, and it is better to know that before the order.