Find out why your PC slows down.
A free online stress test for your PC, with no download. It measures what it holds under load and names the cause.
In short
What this stress test tells you.
SystemCheck is a free stress test that runs in your browser. It loads the processor, memory and graphics hardware for a fixed length of time and records what they deliver from the first second to the last. The result is not one peak score but a curve for each stage, and the report reads the shape of that curve: whether the machine held its speed, stepped down part way, or slid steadily as the run went on.
The CPU stage loads every logical core through Web Workers and attributes each chunk of work to fixed 250 ms windows, so the curve shows what was delivered rather than when a worker happened to report. A short warm-up is excluded, peak is the best one-second mean after it, and decay compares the first seconds after warm-up with the end of the stage. The noise floor is measured on each run, and a shape that does not clear it is not called with confidence.
A browser cannot read temperature, clocks, fan speed or power draw, so none of those appear in a report. A drop in throughput is consistent with several causes, such as a power budget ending, cooling falling behind, a run on battery or another program on the cores, and the report says which conditions it could detect. The methodology sets out every stage and every limit, and the guide to reading results explains when a difference between two runs is real.
The main tests
- Full PC stress testFour stages under sustained load: CPU, memory, GPU and path tracing, with the curve read for you.
- CPU stress testEvery logical core loaded, throughput recorded in 250 ms windows, decay measured against the start.
- GPU stress testA fixed scene at a fixed resolution, every frame timed until the graphics chip has finished it.
- RAM speed testRead and write bandwidth on one thread and on all threads, and latency to cache and main memory.
- Laptop stress testThe same run, read for what laptops do: a boost window that ends and cooling that falls behind.
- PC performance testA quick look at what the whole machine delivers to a browser tab, in plain words.
- Bottleneck testProcessor and graphics measured separately so you can see which one sets the pace.
- GPU benchmarkFrame rate and frame pacing on the same camera path on every machine, so runs compare.
Find your way around
- Every testDisplay, input, audio, camera, network and hardware checks, all in the browser.
- Fix a problemStart from the symptom and get the causes ranked, each with its check.
- Diagnose a symptomSlows after a few minutes, drops frames over time, hot laptop, slow on battery.
- GuidesHow to read a curve, sustained against peak, and when a difference is real.
- ProcessorsThe processor catalogue, with every unverified specification marked as such.
- Graphics cardsThe graphics catalogue, in the same format and with the same rule.
- GlossaryThrottling, frame pacing, power limits and the other words a report uses.
- MethodologyEach stage in full, the noise floor, and every limit stated up front.
Questions people ask
What does SystemCheck test?
It loads your processor, memory and graphics hardware from a browser tab for a fixed length of time and records what they deliver the whole way through. A standard run has four stages: sustained CPU load, memory bandwidth and latency, GPU raster at a fixed resolution, and path tracing. The report reads the shape of each curve, not only its best moment.
How long does a run take?
A standard run takes 3 minutes. A deep run takes 9 minutes and adds display refresh and frame pacing, a combined CPU and GPU load, a soft-shadow stage, and a repeat of the CPU stage on the already-warmed machine.
Why measure sustained performance instead of a peak score?
A peak is what a machine holds for a few seconds. Many laptops boost hard at first and then settle lower once a power budget runs out or cooling falls behind, and the settled level is what you live with. The test compares the start of each stage with its end and reports the difference as decay.
Can a browser read my temperatures or clock speeds?
No. A web page has no access to temperature, clock, fan, voltage or power sensors, so SystemCheck never reports them. It measures delivered throughput and frame times, and a shape in that data is described as consistent with a cause, alongside its look-alikes, rather than stated as a reading.
Is it free, and do I need to install anything?
The test and every measured number in the report are free, with no install and no account. A one-time payment unlocks the diagnosis section of a report. Nothing that was measured is placed behind that payment.
Why do two runs on the same machine differ?
Battery against mains power, a warm machine, and other programs sharing the cores all move the result. The report opens with the conditions that would make a run less trustworthy, such as a hidden tab or a run on battery, and a deep run takes the median of three CPU passes.
The diagnosis
Not how much it slowed. Why. The shape of the curve is the diagnosis, and the report reads it.
A score says the machine got slower. The curve behind it says what happened. Three shapes matter, each with a different cause.
A single step down
A power budget expiring. The machine held its boost for as long as it was allowed to, then dropped to the sustained limit in one move.
A steady slide
Cooling falling behind. Every second under load, a little more is lost, and it does not settle until the fan can keep pace.
A sawtooth
A machine cycling between the two: it cools, boosts, hits the limit, drops, and repeats.
Four stages, then a name for what happened. Does the machine hold what its spec sheet promised, and if not, why not?
Stage 1
Every core, for 60 seconds. The number that matters is the one it settles at.
Parallel Web Workers saturate every logical core and the main thread samples throughput four times a second. Peak arrives in the first second; the rate the machine holds once cooling becomes the limit is what gets reported.
Illustrative run. Your own cores draw their own.
Stage 2
One thread, then six. How much more the memory system gives when asked.
Workers stream 24 MiB arrays to measure bandwidth from one thread and then from all of them, and a pointer chase through a shuffled array measures how long a round trip to memory takes. A machine whose bandwidth barely scales with threads is telling you about its memory channels, not its cores.
One thread
Six threads
Illustrative figures. Not a measurement.
Stage 3
One fixed scene, the same on every card. Frame time, pacing, and how far it falls.
A lit 3D model renders at a fixed 1600 by 900 on a camera path that repeats every ten seconds, and every frame is timed until the graphics chip has finished it. Frames per second, how steadily they arrive, and how far the rate falls are reported. Stutter and slowness look different here, which is the point.
Illustrative frame rates. Not a measurement.
Stage 4
The same scene, soft edges and hard. What realistic shadows cost your card.
The Deep run draws one scene at 1600 by 900 with percentage-closer soft shadows and with plain hard shadows, switching every nine seconds, and times every frame. The gap is the share of frame rate soft shadows take. It is reported, never scored, and the same test runs on its own page.
Illustrative frame rates. Not a measurement.
Stage 5
Samples back to back, each one timed. So the cost of each sample is the measurement.
A WebGL2 path tracer runs samples back to back at a fixed 960 by 540, accumulating each into a float texture, and times every one until the graphics chip has finished it. Samples per second and delivery consistency are reported; a card that delivers unevenly is distinguishable from one that is merely slow.
Illustrative progress. Not a measurement.
Report
Not how much it slowed. Why. The shape of the curve is the diagnosis.
A plateau that steps down once is a power budget expiring. A steady slide is cooling falling behind. A sawtooth is a machine cycling between the two. The report reads the shape, states its confidence, and names the cause with the number to beat on the retest. The next run on the same machine is shown against this one.
Plateau, then a single step down.
Throughput held for 18 seconds, then fell in one transition and stayed there. One fall accounts for 98% of everything the run gave up.
Illustrative shape. Your run draws its own, and the report reads that one.
The limits, stated plainly
What a browser cannot see.
No web page can read a temperature. A cooling-limited machine still shows it, as decay, and decay needs no sensor.
Illustrative shapes. One machine holds near 96% of its peak; one settles at 64%.
Trusted by people who push hardware. Builders, integrators, repair techs.
“I was about to buy a second-hand ThinkPad for 400 quid. Ran the test in the seller's kitchen, watched the CPU stage lose 38 percent in three minutes, and walked away. Best three minutes I spent that week.”
Marcus R. · Buying a used laptop“I hand out one link to clients now instead of a download. They run it, forward me the report, and I know before I drive over whether the cooling is the problem.”
Priya S. · System integrator“My frames were fine for ten minutes and then fell off a cliff every night. The graphics stage showed the same drop at the same point, and the report said thermal ceiling in plain words. New paste, problem gone.”
Devon L. · Gamer, fixed his FPS drops“Exports kept getting slower the longer I worked. I assumed it was the software. The memory stage showed the laptop was fine on mains and lost a third of its speed on battery. Now I edit plugged in.”
Ana M. · Video editor“It says up front what a browser cannot see, and then measures the rest properly. That honesty is why I trust the number it gives me over a spec sheet.”
Tom K. · Sceptical by trade“Forty refurbished desktops in an afternoon, one link each, same test on every machine. The three that graded D all had dust-packed fans. The report told me before the screwdriver did.”
Jordan P. · IT refurbisher“Watching ops per second sink live while the fans spun up taught me more about my machine than any single score ever did. The retest target told me exactly what to aim for after the fix.”
Sana D. · Curious about her own PC“New build, new cooler, and I wanted proof it was worth it. Before and after in two runs: decay went from 31 percent to 6. That is the number I put in the forum post.”
Leo V. · First-time PC builder“Twelve laptops in the classroom, three complaints a week that they were slow. Ran the test on all of them from the same link. Two were throttling badly, ten were fine. Two repairs instead of twelve.”
Ravi N. · School IT admin“I asked in a forum why my new PC felt slower than the reviews said. Someone sent this. Three minutes later I could see it was hitting a power limit under load, and I knew what to change in the BIOS.”
Emily W. · New PC underperforming“Average FPS looked great and the racing still felt wrong. The consistency figure caught frame pacing my overlay was hiding. Once I could see it, I could fix it.”
Chris B. · Sim racer“Customers say slow, I need a number. I run it in the browser while they watch, the report says what is happening in plain words, and the repair conversation is over in a minute.”
Nadia F. · Repair shop ownerEvery part, with its specifications. And what they do not tell you.
1,737 processors and every graphics card we can source properly, with side-by-side comparisons. 218 carry a full sheet today; the rest say “not verified” rather than guess.
Fix a problem
Start from the symptom. Causes ranked, each with its check.
- Why is my laptop so slow?
- Why is my laptop overheating?
- What is thermal throttling, and is my laptop doing it?
- Why does my FPS drop after 10 minutes of gaming?
- Why is my laptop fan so loud?
- Why is my laptop battery draining so fast?
- Why is my PC suddenly slow?
- Why is my laptop slower on battery?
- How do I check if my Windows laptop is thermal throttling?
- Is my laptop dying?
- What should I check when buying a used laptop?
- How do I stress test my laptop?
Writing
Better as you read. Guides and notes on measuring honestly.
- Memory latency vs bandwidth: what a random-walk test showsBandwidth is how many bytes memory moves per second; latency is how long one dependent load waits. How SystemCheck measures each and what they cannot say.
- Single-thread vs all-thread memory bandwidthWhy SystemCheck measures memory bandwidth on one thread and then on several, what the ratio between the two shows, and what it cannot prove.
- What "not measured" means in your reportWhen a stage cannot run or its data cannot be trusted, SystemCheck says not measured and gives the reason instead of estimating a number.
- Reading the confidence interval on every figureEach headline figure in a SystemCheck report carries an interval: the spread inside that one run. How it is computed, what it covers and what it does not.
- Soft shadows as a GPU load: the A/B methodHow SystemCheck measures what soft shadows cost a GPU by alternating soft and hard segments in ABBA order, so that warming up does not flatter either one.
- What a measurement revision isEvery SystemCheck report is stamped with a measurement revision. What changes one, why reports across some revisions are not compared, and how to read it.