The institutional code of ethics governing RDPChoice research engineers, benchmark data integrity, mystery procurement procedures, and provider corrections.
At RDPChoice, our technical integrity is our sole asset. We enforce a strict institutional firewall between our systems benchmarking laboratory and our commercial operations. Benchmark analysts and research engineers do not participate in commercial sponsorship negotiations, and commercial personnel have no access to modify, adjust, or delay the publication of raw hardware benchmark results.
When our laboratory identifies severe performance throttling—such as an advertised NVMe instance dropping to 40 IOPS under sustained load or a Forex VPS exhibiting 90ms packet jitter during market open—we publish the findings prominently, regardless of commercial relationships.
Hosting vendors frequently offer “evaluation units” to tech reviewers. We refuse all such offers. When vendors know an account belongs to a benchmark lab, they route that VM to an unloaded hypervisor with guaranteed dedicated resources. By purchasing every instance anonymously using rotating commercial cards, we experience the identical noisy-neighbor conditions and provisioning delays faced by everyday retail customers.
No review is written based on a 15-minute speed test. Every reviewed instance runs continuous monitoring scripts for a minimum of 72 hours (and up to 720 hours for low-latency nodes) to evaluate long-term thermal throttling, memory leaks, and hypervisor oversubscription.
Every review includes the exact synthetic test commands used (FIO configurations, Sysbench flags, traceroute parameters). If our results differ from a provider’s marketing claims, we provide full command logs so third-party engineers can replicate our findings.
We believe in rigorous fairness. When a hosting provider upgrades its underlying hardware cluster (e.g. migrating from Intel Xeon E5 to AMD EPYC Genoa) or identifies a factual error in our review, we provide an expedited right-of-reply mechanism: