We’re excited to announce that the locations for the new, state-of-the-art Brixly EU and US clouds have been finalised and will soon be operational!

As many of you know, since Enix acquired Brixly, one of our top priorities has been deploying hardware in the EU (Germany) and US. Currently, Brixly rents servers from well-known providers in these regions (who shall remain unnamed, though you can probably figure out who they are). Over the past 12 months, this arrangement has led to numerous challenges. The primary concern for Enix has been our lack of control, leaving us at the mercy of these providers. For example, in recent months, these providers have unnecessarily null-routed servers during minor DDoS attacks, resulting in complete outages for our customers. Additionally, the rented servers are nearing capacity, and further upgrades would pose significant challenges.

As I’ve mentioned before, Enix's goal is to own the entire infrastructure stack from top to bottom—and that’s exactly what we’re doing. We will be deploying high-spec servers to both locations, set up in a fully redundant cluster. Each server will feature AMD Epyc CPUs, high-performance DDR4 memory, and enterprise-grade Micron NVMe storage. This is part of Enix’s significant investment in our next-generation cloud infrastructure in the UK, EU, and US.

Of course, high-spec servers need to be housed in top-tier data centres, so we’ve chosen the gold standard. In Frankfurt, our servers will be hosted at Equinix FR5, ensuring the best possible connectivity in Central Europe. In the US, we’ve selected CoreSite VA1 in Virginia, located near Dulles Airport, just 30 minutes outside Washington, D.C.

Additionally, we’ve partnered with Voxility to provide DDoS-protected IP transit. Voxility has consistently delivered exceptional performance for Brixly's UK transit, making it the clear choice for extending this partnership to our EU and US locations. Not only does Voxility offer top-tier DDoS mitigation, but it also provides the very best in low-latency connectivity worldwide, ensuring fast and reliable service across the globe.

Again, keeping with our ‘own the stack’ philosophy, Enix is a RIPE member, meaning we own, and will always own the IP addresses used on the new transit permanently.

And finally, no announcement would be complete without a few hardware photos! Here are a couple of shots, personally taken, of the very servers going into these locations.

The AMD Epyc Milan CPU flanked by 1TB of DDR4 memory. Don’t worry, the heatsink was added after the photo. Towards the back you can see two internal m.2 NVMe drives for the OS poking out from behind the memory bank on the left. The chassis itself is the latest generation Asus 1U rack server with redundant PSU.

A tray of 15.36TB Micron 7450 Pro NVMe SSDs ready to go into the nodes. Each node can hold up to 12 and each drive has dedicated PCIe access to the CPU. In total we’ll deploy 1.5PB of raw storage in our clouds, with triple data replication for the very ultimate in protection, and seamless capacity expansion.

We aim to have the clouds online in 2-4 weeks. All customers will, of course, be given plenty of notice of any migrations, and we will do our best to make the migrations as flexible and smooth as possible.

Elliot