Difference between SeqCst and Release performance

use std::sync::atomic::{AtomicUsize, Ordering};
use std::thread;

fn main() {
    static COUNTER: AtomicUsize = AtomicUsize::new(0);

    let cores = core_affinity::get_core_ids().expect("Failed to get core IDs");

    let writer_core = cores.first().unwrap().clone();
    let reader_core = cores.last().unwrap().clone();

    let writer = thread::spawn(move || {
        core_affinity::set_for_current(writer_core);
        for i in 1..=1_000_000_000 {
            COUNTER.store(i, Ordering::Release);
        }
    });

    let reader = thread::spawn(move || {
        core_affinity::set_for_current(reader_core);
        loop {
            let value = COUNTER.load(Ordering::Acquire);
            if value == 1_000_000_000 {
                break;
            }
        }
        println!("Counter reached 1 billion!");
    });

    writer.join().unwrap();
    reader.join().unwrap();
}