Work-Stealing Architectures

Optimize load balancing in multi-threaded runtime environments using thread-local double-ended queues.