Live server performance, side by side
These numbers come from the servers our customers are using right now, and they refresh every hour. They show how the servers actually run, not the most the hardware could do.
The plans are built for different jobs: they differ in disk type, server size and how many users share a server. That is why the trophies compare HDD plans with HDD plans and SSD plans with SSD plans, never across the two.
Bold is the average across the plan's servers; the grey line under it is the median, the server in the middle. marks the best average in each row, HDD plans and SSD plans judged separately: more storage, more drives and faster measured reads win; on the other rows lower wins. Swipe the table sideways to see every plan.
| Live, per server | HDD plans | SSD plans | ||||
|---|---|---|---|---|---|---|
| Dragon-R | M10G | M1000 | V10G | M10G SSD | M1000 SSD | |
| Users per server | 8.7median 7 | 26.5median 26.5 | 26.4median 20 | 8.9median 7 | 33.9median 31 | 32.5median 30.5 |
| Network port (plan spec) | 20 Gbps | 10 Gbps | 1 Gbps | 10 Gbps | 10 Gbps | 1 Gbps |
| Storage per server | 30.4 TiBmedian 29 TiB | 71.3 TiBmedian 58 TiB | 41.6 TiBmedian 38 TiB | 31.2 TiBmedian 32.6 TiB | 9.4 TiBmedian 7.2 TiB | 5.4 TiBmedian 5.4 TiB |
| Drives per server | 4.7median 4 | 6.8median 6 | 4.8median 4 | 4.8median 4 | 4median 4 | 2.1median 2 |
| Disk response time | 29.3 msmedian 24.8 ms | 39.5 msmedian 31.4 ms | 31.7 msmedian 29.2 ms | 16.7 msmedian 17.2 ms | 5.7 msmedian 4.1 ms | 2.5 msmedian 2 ms |
| Waiting on disk | 19.6%median 17.3% | 21.1%median 17.5% | 22.2%median 17.6% | 6.3%median 5% | 28%median 31.3% | 8.9%median 1.3% |
| Disks busy | 23.1%median 24.3% | 26%median 25.2% | 23%median 20.3% | 10.8%median 7.9% | 14.6%median 14.2% | 5.6%median 4.3% |
| Measured read speed | 405 MB/smedian 433 MB/s | 605 MB/smedian 584 MB/s | 241 MB/smedian 331 MB/s | 317 MB/smedian 346 MB/s | 785 MB/smedian 766 MB/s | 861 MB/smedian 877 MB/s |
Measured read speed comes from a read-only test of the disk array behind each server (large sequential reads), run on a sample of each plan's servers while they were quiet. The server reads at least this fast. Higher wins on this row; each figure comes from a sample of the plan's servers, not from every one.
Updated 24 Oct 2026, 14:37 UTC, refreshed hourly. Raw data (JSON).
What the numbers mean
- Users per server: how many accounts share one server. Every account type counts.
- Storage per server: the total disk space of one server, in TiB.
- Drives per server: how many physical drives hold that space. For servers that run as virtual machines this is the drives in the array behind the machine, not the virtual disks it sees.
- Disk response time: how long a disk read takes to complete, averaged over the past week.
- Waiting on disk: how much of the week every active process on the server was waiting for the disks, averaged over the servers that recorded any waiting. Lower means less waiting.
- Disks busy: how much of the week the disks had work to do. This is time, not capacity used.
These are live readings of real use, not the maximum each server can do. They cover the past week, and none of them is a promise for the next one. A server here means the machine or virtual machine a seedbox or storage box runs on; Storage Boxes run on the M10G servers.