Hardware Atomic Operations: `Interlocked` Class & Lock-Free Counters

1Concept

`System.Threading.Interlocked` utilizes CPU hardware lock bus prefix instructions (`LOCK XADD`, `LOCK CMPXCHG`) to perform atomic increments and compare-exchange operations in a single CPU cycle without operating system lock kernel transitions.

2Architecture Diagram

Thread 1 (Core 1) ──┐
                     ├──> [ Hardware Bus LOCK CMPXCHG ] ──> Target Integer in RAM
Thread 2 (Core 2) ──┘    (Guaranteed Atomic Update without OS Locks!)

3Code Example

using System;
using System.Threading;
using System.Threading.Tasks;

public class InterlockedDemo
{
    private static long _requestCount = 0;

    public static void Main()
    {
        Parallel.For(0, 100_000, i =>
        {
            // Hardware lock-free atomic increment
            Interlocked.Increment(ref _requestCount);
        });

        Console.WriteLine($"Final Atomic Request Count: {_requestCount}");
    }
}

4Expected Output

Final Atomic Request Count: 100000

5Key Takeaways

• ✓Use `Interlocked.Increment`, `Decrement`, `Add`, and `CompareExchange` for high-performance lock-free counters.
• ✓`Interlocked` is ~20-50x faster than standard `lock(obj)` blocks.
• ✓`CompareExchange` enables lock-free optimistic spin loops.