Performance is
science, not opinion.
Every recommendation on this site comes from a real test, a real benchmark, and a real rig. No affiliate-first rankings. No spec-sheet copy-paste. Just 17 years of IT experience turned into numbers you can actually use.
IT Manager by day.
Performance obsessive by night.
I’m Kevin — a 31-year-old IT professional with 17 years of hands-on experience managing infrastructure, diagnosing bottlenecks, and squeezing performance out of hardware that was never supposed to perform that well.
I started PerfGamer because I kept seeing the same thing: gaming guides written by people who had never run a benchmark in their lives. Affiliate links dressed up as recommendations. Marketing language recycled as advice. Coolers ranked by box art.
This site is the opposite of that. Everything I write comes from something I’ve built, tested, measured, or broken. When I tell you the Thermalright PA 120 SE is the best $35 you can spend, it’s because I’ve tested it against units twice the price and logged the delta-T myself.
My day job means I think in systems — not just individual parts. A PC isn’t a collection of specs; it’s a thermal loop, a power delivery chain, a latency stack. That’s the lens I bring to every article.
The lab process
Every article starts with a question, not a conclusion. Here’s what the process looks like from idea to published guide.
Start with a specific claim worth testing. “Does disabling HPET improve 0.1% lows?” — that’s a testable question.
Control variables, establish a stable baseline, document hardware config. Reproducibility is the goal.
Run tests under consistent conditions. Capture frame times, temps, acoustics — not just average FPS.
Write what the data says — even when it contradicts popular advice. Especially then.
Hardware in the lab — Tests run on an AM5 platform (Ryzen 7 9800X3D), RTX 4080, 32GB DDR5 6000 CL30, and multiple case configurations. Thermal data captured with FLIR and logged via HWiNFO. Frame timing via CapFrameX.
Five things that don’t change.
A guide that took three weeks to test is worth more than one that took three hours to write. This site publishes less, but publishes better.
Every claim comes with methodology. If you can’t replicate it, it’s not a finding — it’s a guess. I document platform, workload, and conditions.
Most people building a gaming PC have $500–$1,000 to work with. That’s the lens. Premium picks appear when they genuinely earn their price — not because they pay higher affiliate rates.
If a tweak makes no measurable difference, I’ll say so. If the $35 cooler beats the $150 one in real gaming loads, the article says that — even if the $150 one has a better affiliate program.
Performance isn’t mystical. There’s a reason CPUs thermal-throttle, a reason frame times stutter, a reason your GPU underperforms. The answers are in the data — this site exists to find them and explain them clearly.
How PerfGamer makes money
Affiliate links: Some links on this site use affiliate programs (Amazon Associates and others). If you buy something through a link here, I may earn a small commission — at no extra cost to you. This is how the site covers its costs.
What it doesn’t change: Affiliate relationships never dictate recommendations. If a product underperforms in testing, it doesn’t appear in a top pick — regardless of commission rate. The Thermalright PA 120 SE earns a few cents per sale. It’s still the first cooler I recommend, because the data says so.
No sponsored reviews: Manufacturers don’t pay for coverage. I buy most hardware myself or return review units. No one has editorial input except the benchmark results.