Transparency

Our Hardware

We think you should know exactly what you're running on. Here's a node-by-node look at the real data center hardware behind RackGenius, from game nodes to VPS hosts to bare metal. Every CPU, memory and storage figure below describes a machine actually racked in Detroit or Grand Rapids, not a spec sheet we aspire to.

Hardware varies as we grow. On shared services like VPS and game hosting we can't promise a specific chip, so your service may land on any host in its tier. Dedicated servers are the exception, where you get the exact machine listed.

Loading fleet

Preparing the hardware list

Hardware FAQ

FrequentlyAskedQuestions

  • On dedicated servers, yes. The machine listed is the one you receive. On shared services like VPS and game hosting, your service runs on one of the hosts in its tier, so the exact chip can vary while the performance class stays the same.
  • It syncs automatically from our internal inventory once a week, so retired machines drop off and new ones appear on their own. The "Fleet data updated" date near the top shows the last sync.
  • ECC is server memory that detects and corrects bit errors on the fly, which protects against silent data corruption. On the newer consumer platforms that do not take ECC DIMMs, the CPUs still use on-die ECC where supported.
  • Value is high core count and large RAM at a lower price, great for background and bulk workloads. Ryzen and Premium lean on high-clock desktop-class chips for latency-sensitive work. LTO stands for Limited Time Only, a rotating batch we offer while stock lasts.
  • Older dual-socket platforms still deliver a lot of cores, memory bandwidth, and capacity for the money, which is exactly what our Value and LTO tiers are built around. Latency-sensitive game and premium nodes use high-clock Ryzen instead.
  • Everything lives in Michigan, on-net at our Detroit (DTW) and Grand Rapids (GRR) data centers. Use the location switcher to see the fleet at each site.
  • Each card shows the CPU model and core count, the memory type and capacity, and the storage layout, along with badges for ECC memory and redundant power. Those are the data center hardware components that actually decide how a service performs, so they are the ones we publish instead of a vague tier name.
  • We match the infrastructure servers in a tier to the work they do. Value and LTO nodes use dual-socket platforms for core count and memory capacity per dollar. Ryzen and Premium nodes use high-clock chips where single-thread speed decides the outcome. Storage is NVMe anywhere latency matters, with SATA SSD kept for the tiers where capacity is worth more than throughput.
  • Yes. If you need a specific CPU, more RAM, extra storage, or additional IPs, open a ticket and we will build a machine to spec.

Questions? Let's chat.

Get in touch with our support team

Email Support

Subscribe to our Newsletter

Get updates, offers, and news straight to your inbox.