FPU

What this benchmark does

This benchmark renders an ASCII Mandelbrot set repeatedly. For each iteration it walks a fixed 78 by 78 grid of complex-plane sample points, computes the Mandelbrot escape iteration count for each point, and writes either * or a space depending on whether the point stays in the set.

The benchmark therefore mainly measures floating-point arithmetic, tightly nested loops, branch-heavy escape checks, and repeated text output while keeping the rendered image identical across languages.

Benchmark system

• Apple M1 Pro (arm64)
• 10 CPU cores

Use ../../run -b fpu to measure this benchmark. It pre-builds compiled implementations and measures each benchmark command with /usr/bin/time.

Results