Run a short single-core and multi-core processor benchmark and get a clear browser-based CPU score.
The workload runs locally in browser workers. It does not reveal processor temperature, clock speed or the exact CPU model.
The benchmark runs a fixed mix of integer multiplication, bitwise operations and square roots in Web Workers. One pass measures a single worker for about 900 milliseconds; another uses up to eight workers for about 1,050 milliseconds. Completed inner-loop iterations per second and the published formula produce a compact browser score.
The result measures how this browser executes this particular workload at this moment. CPU architecture, clock behavior, browser engine, worker scheduling, power mode, thermal limits and background activity all contribute. Repeat under controlled conditions before comparing a change.
This reflects one worker's JavaScript workload iterations per second and is often relevant to web tasks that cannot spread evenly across cores.
This reflects combined workload iterations across the browser-selected worker count, capped at eight to keep the test short and comparable.
The rounded score equals single-worker iterations per second divided by 100,000 plus parallel iterations per second divided by 250,000.
A small change is normal. Sustained large differences under the same conditions can point to power, heat or competing workloads.
This is not a comprehensive processor benchmark. It does not measure clock speed, temperature, energy use, native-code performance, every instruction type or application performance, and it should not be compared directly with scores from other benchmark products.
Read our complete testing methodology →It is a real timed JavaScript workload executed in isolated browser workers. It is useful for comparing the same device and browser, but it is not a replacement for a native hardware benchmark.
No. Browsers expose the number of available logical processors but normally hide the exact CPU model, clock speed and temperature for privacy.
Temperature, battery mode, background apps, browser version and power management can all change short benchmark results.
It means higher throughput in this short browser workload. Real applications use different code, graphics, memory and storage, so their ranking can differ.
The cap limits heat and keeps the short test more comparable on machines with very different logical-thread counts.
The workload is brief and uses normal browser code, but any demanding task can raise activity temporarily. Stop if the device behaves abnormally and avoid repeated continuous runs.