Machine Language: Count Faster on 6502

I’ve done a lot of silly math, ciphers and asked a lot of my vintage hardware on this site over the years. But I’ve not talked a ton about optimizing code for faster results.

Today, we will count from 0 to 2^24-1 (16,777,215). Nothing else. Count, time it, and see how fast we can get it to go.

Rust

In Rust on my 2020 MacBook Pro:

use std::time::{Instant};

fn main() {
    // Start timing
    let start = Instant::now();
    // Loop from 0 to 16777215
    for _i in 0..=16777215 {
        // The loop body is empty, just like in the BASIC code
    }
    // Calculate elapsed time
    let duration = start.elapsed();
    // Convert elapsed time to seconds
    let elapsed_seconds = duration.as_secs() as f64
        + (duration.subsec_nanos() as f64 / 1_000_000_000.0);
    println!("Counting to 16777215 took {} seconds", elapsed_seconds);
    println!("That's {} additions per second", 16777216.0 / elapsed_seconds);
    println!("On 2 GHz Quad-Core Intel Core i5")
}

Counting to 2^24 in Rust