Skip to main content

PassMark PerformanceTest

PassMark PerformanceTest benchmarks CPU, memory, disk, and graphics subsystems, aggregating them into comparable Mark scores backed by a large public database. Component sub-scores help diagnose system bottlenecks.

PassMark PerformanceTest is a broad whole-system PC benchmark that evaluates the major hardware subsystems and combines them into easy-to-compare scores. It is widely used for procurement, troubleshooting, and comparison because its large public database — the PassMark CPU and GPU charts — provides ranking context for almost every consumer and server processor and graphics card.

The suite runs separate test groups covering CPU (integer, floating-point, compression, encryption, sorting, physics), memory (read, write, latency, threaded access), disk (sequential and random read/write across storage devices), 2D graphics (vector, fonts, GUI), and 3D graphics (DirectX rendering tests). Each group yields a sub-score, and the tool aggregates them into an overall system rating.

What It Measures

The headline figures are the CPU Mark, Memory Mark, Disk Mark, 2D Mark, and 3D Mark sub-scores, plus an overall PassMark Rating that blends them. These are relative numbers calibrated so users can rank a machine against others. The CPU Mark in particular is one of the most cited cross-processor comparison figures because of PassMark's extensive submitted-results database.

Methodology

Each subtest runs a defined workload and converts its raw result (operations per second, MB/s, frames per second) into a normalized score scaled against reference hardware. The suite can run single-threaded and multi-threaded CPU tests to show both per-core and aggregate capability. Users can upload results to the online database, which aggregates thousands of submissions into the published charts. Configuration is recorded with each result.

How to Interpret Results

Use the component sub-scores to find bottlenecks: a strong CPU Mark with a weak Disk Mark points to storage as the limiter. The overall rating suits quick whole-system comparison, but component scores are more diagnostic. For cross-machine comparison, prefer the online charts, which average many submissions and reduce single-run noise. Distinguish single-thread from multi-thread CPU Marks depending on whether your workload is parallel.

Limitations

PassMark scores are synthetic and relative, so they indicate comparative standing rather than performance in a specific application. Submitted-database figures can include overclocked or misconfigured systems, adding noise. The 3D tests are lightweight compared to dedicated graphics benchmarks, and the suite does not capture sustained thermal behavior or real workload mixes. As with all benchmarks, the numbers are a starting point, not a substitute for testing your actual software.