Receipts
Every claim.
Something behind it.
Benchmarks, the capacity policy, uptime and credits, the upgrade policy. Every one linked here, open to inspection.
Independent benchmarksQuarterly third-party trials since January 2025, and five competitors in one suite. Every result linked.
Stock ControlHow we cap every node at 60% load, and how you can check yours.
Uptime and creditsThe 99.9% SLA over any 30-day period. Miss it and the credit is issued without being asked.
Level UpDated spec upgrades. Same plan code, same price.
Changes and incidentsWhat happened, when, and what we changed afterwards.
Awards and reviewsThird-party awards and the Trustpilot score, linked to the source.
Benchmarks
Third-party runs, linked to every result.
Five competitors, the same test suite, and a link to each run on VPSBenchmarks. Nothing is cherry-picked, and you can rerun it on your own server.
How to read it
Single core is one thread's speed. Multi core is all of them together. Storage is fio at 4k and 64k blocks. Network is iperf3.
The run
2,323 Geekbench 6 single-core on the 6 GB Performance VPS, run on 2025-04-22. Open the run.
Geekbench - CPU Speed Test
Geekbench 6 numbers, Single Core & Multi Core
- GoDaddy - Managed VPS – 32GB 8vCPU
- Hetzner - CAX21 – 8GB 4vCPU
- RackNerd - 4.5GB KVM VPS – 4.5GB 3vCPU
- Contabo - SSD1 – 6GB 4vCPU
- OVHcloud - c3-8 – 8GB 4vCPU
- AMD EPYC KVM 6G – 6GB 3vCPU
Storage Speed - Fio 4k
Combined random read and write test (50/50). Fio measured numbers for 4k block size.
- GoDaddy - Managed VPS – 32GB 8vCPU
- Hetzner - CAX21 – 8GB 4vCPU
- RackNerd - 4.5GB KVM VPS – 4.5GB 3vCPU
- Contabo - SSD1 – 6GB 4vCPU
- OVHcloud - c3-8 – 8GB 4vCPU
- AMD EPYC KVM 6G – 6GB 3vCPU
Storage Speed - Fio 64k
Combined random read and write test (50/50). Fio measured numbers for 64k block size.
- GoDaddy - Managed VPS – 32GB 8vCPU
- Hetzner - CAX21 – 8GB 4vCPU
- RackNerd - 4.5GB KVM VPS – 4.5GB 3vCPU
- Contabo - SSD1 – 6GB 4vCPU
- OVHcloud - c3-8 – 8GB 4vCPU
- AMD EPYC KVM 6G – 6GB 3vCPU
Network Speed - iperf3 ipv4
Iperf3 network transfers speeds. For each tested VPS only the fastest transfer for each direction (download and upload) in those charts.
- GoDaddy - Managed VPS – 32GB 8vCPU
- Hetzner - CAX21 – 8GB 4vCPU
- RackNerd - 4.5GB KVM VPS – 4.5GB 3vCPU
- Contabo - SSD1 – 6GB 4vCPU
- OVHcloud - c3-8 – 8GB 4vCPU
- AMD EPYC KVM 6G – 6GB 3vCPU
Reproduce it
One command on any Linux server:
curl -sL https://yabs.sh | bash
The YABS guide explains each number.
Where it runs
The same nodes every plan lands on. See VPS plans.
Quarterly trials, every grade as given.
VPSBenchmarks buys a plan each quarter, runs its own suite, and grades five categories. 8 trials since January 2025, including the bad grades. Checked 2026-09-09 21:11 UTC against the hoster page.
| Date | Plan tested | CPU | Paid | Score | Grades | Run |
|---|---|---|---|---|---|---|
| 2026-09-07 | P10 Performance VPS | 1 vCPU AMD EPYC 4565P | $8.00/mo | 83 (65 raw) | B D A A Eweb cpu stab disk net | Open → |
| 2026-06-30 | Core 4X VPS | 4 vCPU AMD EPYC 4565P | $29.99/mo | 87 (85 raw) | A A B A Bweb cpu stab disk net | Open → |
| 2026-03-29 | Pro 5X VPS | 5 vCPU AMD EPYC 4565P | $36.99/mo | 86 (86 raw) | A A A A Cweb cpu stab disk net | Open → |
| 2026-01-02 | Nano 1X VPS | 1 vCPU AMD EPYC 4565P | $4.99/mo | 90 (72 raw) | B D A A Cweb cpu stab disk net | Open → |
| 2025-10-07 | Max 7X VPS | 7 vCPU AMD EPYC 4565P | $54.99/mo | 80 (87 raw) | A A A A Bweb cpu stab disk net | Open → |
| 2025-07-14 | Micro 2X VPS | 2 vCPU AMD EPYC 4564P | $14.99/mo | 77 (60 raw) | B C B C Dweb cpu stab disk net | Open → |
| 2025-04-22 | AMD EPYC KVM 6G rev1 | 3 vCPU AMD EPYC 4564P | $31.99/mo | 68 (65 raw) | B B C B Cweb cpu stab disk net | Open → |
| 2025-01-30 | AMD EPYC KVM 12G rev1 | 4 vCPU AMD EPYC 4564P | $49.99/mo | 59 (68 raw) | A B C B Dweb cpu stab disk net | Open → |
The score is VPSBenchmarks' overall grade weighted by price; the raw score is in parentheses. Plan names are what VPSBenchmarks recorded at the time; the current names are on the product pages.
Stock Control
We stop selling on a node at 60% average CPU load.
The other 40% is headroom for your spikes. Nobody else's load slows yours.
- The ruleNo new server lands on a node once its average CPU load reaches 60%. Existing servers keep the headroom.
- What it protectsSingle-core speed under load, which is what a busy game server or web app feels first.
- Check yoursRun
vmstat 1 10on your server and read the last column, steal time. Single digits under load are fine. The YABS guide walks through it. - The readingsEvery node's CPU, RAM, and disk use, as the status monitor reported them at the last refresh. The orange mark on each CPU bar is the 60% cap.
| Node | CPU | RAM | Disk | 30-day uptime |
|---|---|---|---|---|
| honest-husky | 1.2% | 36.9% | 12.6% | 100.00% |
| live-quail | 2.1% | 35.1% | 14.5% | 100.00% |
| loved-koala | 5.1% | 51.0% | 6.3% | 100.00% |
| moral-seal | 0.3% | 19.7% | 9.5% | 100.00% |
| prompt-bream | 6.7% | 38.3% | 13.9% | 100.00% |
| smart-moray | 0.1% | 14.8% | 4.7% | 100.00% |
| strong-iguana | 4.4% | 41.0% | 24.9% | 100.00% |
| storagenode01 | 0.2% | 74.4% | 35.8% | 99.99% |
| gs042 | 8.0% | 36.3% | 87.6% | 99.99% |
| gs043 | 4.3% | 10.6% | 85.2% | 100.00% |
| gs045 | 0.3% | 65.7% | 30.6% | 100.00% |
Readings from status.nobullnetworks.com at 2026-09-09 22:11 UTC; the live values are there, node by node. The cap is on average CPU; a reading above it for a moment is a spike, not a breach.
Credit-Backed SLA
99.9% network uptime. Miss it and the credit is issued without being asked.
The terms are short and they are public.
- The target99.9% network availability, measured over any 30-day period.
- The creditIssued automatically to your account, capped at 50% of the affected service's monthly rate. No ticket needed.
- The scopeOur network from the redundant spine to our upstream carriers. Your own server, software, or connection is outside it.
- The termsEvery clause is on the SLA and Money-Back Guarantee page.
- Live statusCurrent status and incident history at status.nobullnetworks.com.
- Credit recordCredits issued under the SLA are recorded here, by period, as they happen.
| Monitor | What it watches | 30-day uptime | Now |
|---|---|---|---|
| Transit Lumen - WA1 | Transit carrier | 99.88% | Online since 6d ago |
| Transit WholeSail - WA1 | Transit carrier | 100.00% | Online since 16w ago |
| Transit Zayo - WA1 | Transit carrier | 100.00% | Online since 24w ago |
| WA1 Zone1 IPv4 End-point | Zone endpoint | 100.00% | Online since 26w ago |
| WA1 Zone1 IPv6 End-point | Zone endpoint | 99.98% | Online since 2w ago |
| WA1 Zone2 IPv4 End-point | Zone endpoint | 100.00% | Online since 22w ago |
| WA1 Zone2 IPv6 End-point | Zone endpoint | 99.98% | Online since 2w ago |
| Client Portal | Customer-facing service | 99.98% | Online since 2d ago |
| Game Panel | Customer-facing service | 100.00% | Online since 7w ago |
| NoBull Networks Website | Customer-facing service | 99.75% | Online since 2d ago |
Readings from the status page at 2026-09-09 22:11 UTC. Each monitor links to its full uptime report; per-node readings are under Stock Control above.
Level Up
When the spec goes up, the code and the price do not.
When hardware or margins allow, a plan gains resources. Existing customers get it automatically. Every one is dated here.
| Date | Plan | Change |
|---|---|---|
| 2026-08-09 | Performance VPS P20 | 1 vCore to 2 vCores. Code and price unchanged. |
| 2026-08-09 | Foundational VPS F20 | 1 vCore to 2 vCores. Code and price unchanged. |
Dates come from the change to our published plan data. Plan codes never change; new sizes get new numbers.
Changes and incidents
What happened, and what we changed.
Dated records, written by the engineers involved, including the ones that are not flattering.
| Date | Record | Read |
|---|---|---|
| 2026-09-06 | Support email import failed for 109 days. Outbound mail kept working while inbound tickets did not. Correction of errors, with the monitoring we added. | The write-up → |
| 2026-08-22 | Host Migration. The host has failed for GS042 and we are actively in the progress of migrating it to another node note. | Status page resolved → |
| 2026-08-01 | Game Servers impacted. Some of our customers will note that their game servers currently are unresponsive in the console and they may not be able to be connected to. | Status page resolved → |
| 2026-03-08 | GS043 Server Recovery. gs043 has had a hardware issue we are actively working to recover. | Status page completed → |
| 2026-03-07 | Data Center Migration. During this activity you should expect degraded services to times of service outage (1 hour+) for your services as well as NoBull Networks portals and websites. | Status page in progress → |
| 2026-03-06 | Data center move to Lumen Seattle 1. A full move of customer equipment and data out of a closing facility, announced with the timeline. | The notice → |
Incidents post to the status page as they happen; this list is refreshed from it (last at 2026-09-09 22:11 UTC) and the write-ups join it afterwards.
Third parties
Awards and reviews, linked to the source.
The homepage shows the three most recent awards. Every one stays here, with the Trustpilot score as it stands.
