Optimized use of launched workers in parallelism code

Optimized use of launched workers in parallelism code

Edit