Integer operations
Integer arithmetic, multiplication and bitwise operations
— ms
← Back to BenchLab BENCHLAB / HARDWARE
CPU Benchmark
Measure browser-based CPU performance with integer, floating-point, and parallel workloads. For consistent results, close other heavy apps.
CPU PERFORMANCE SCORE
— points
Run the benchmark to measure your processor's performance.
Ready to test
CPU TEST SUITE
3 TESTS Workload results
Floating-point operations
Repeated numeric calculations and mathematical functions
— ms
Parallel performance
Multiple Web Workers calculating at the same time
— ms
Waiting to start
This measures JavaScript workloads, not every CPU instruction or the processor's full native performance. Parallel results depend on browser scheduling, power settings, and background activity.
ABOUT THIS BENCHMARK
How to read the results
This tool measures three browser-based JavaScript workloads. It is a relative BenchLab comparison, not a universal hardware rating or a complete measurement of native CPU performance.
Three workloads
- Integer operations
- 2.5 million integer arithmetic and bitwise iterations using integer multiplication and shifts.
- Floating-point operations
- 1 million calculations combining
Math.sqrtandMath.sin. - Parallel performance
- Each selected Web Worker performs 2 million integer iterations. You can choose 2, 4, 6, or 8 workers.
Timing and score
The integer and floating-point workloads each run once and are
timed with performance.now(). The parallel timer starts
before workers are created and ends when every selected worker
replies, so worker startup, message coordination, scheduling, and
the worker calculations are included. There is no warm-up or
multi-sample median in this benchmark.
The score is Math.round(1000 × ((100 / integer ms) × (100 / floating-point ms) × (500 / parallel ms))1/3).
Lower elapsed times produce a higher relative score. The factors
100, 100, and 500 are fixed
scoring targets, not certified performance ratings.
Compare fairly
Keep the same browser, device, power mode, and worker count when comparing runs. Close unnecessary tabs and applications, avoid other heavy work, and repeat the test under similar thermal and background conditions.
Browser engine behavior, CPU scheduling, available resources, and other processes can affect the result. Parallel timings are especially sensitive to worker scheduling and coordination, so results from different browsers or devices are not directly interchangeable.