When should I use lock-free vs wait-free algorithms?

What are lock-free and wait-free algorithms?

Lock-free algorithms allow threads to make progress without locking, while wait-free guarantees that every thread will complete its operation in a finite number of steps.

When should I use lock-free algorithms?

Use lock-free algorithms when you need high throughput and can tolerate some threads being delayed.

And when is it better to use wait-free algorithms?

Wait-free algorithms are ideal when you need guaranteed response times for all threads, such as in real-time systems.

Can you give me an example of each?

Sure! A common lock-free example is the Michael-Scott queue, while a wait-free example is the Treiber stack.