Awesome video as always! Is this approach always better than using a condition variable and a queue / linked list? I understand that buffering will likely result in less blocks, but it might result in longer waiting periods, especially for the processing thread. My guess is that if you fine tune and profile your buffer size very well according to the generator/processing execution times it will perform better. Let me know your thoughts!
What a wonderful video! Simple but great example, though we need some C++20 features such as span, binary_semaphore, jthread. Btw, I guess -pthread would be better than the -lpthread flag! Thanks for the great explanation ;)
wonderfully!!! And i Iwant to systematically learn parallel computing, which series should i start with? plz
Awesome video as always! Is this approach always better than using a condition variable and a queue / linked list? I understand that buffering will likely result in less blocks, but it might result in longer waiting periods, especially for the processing thread.
My guess is that if you fine tune and profile your buffer size very well according to the generator/processing execution times it will perform better. Let me know your thoughts!
Linked lists suck for any kind of iterative work like this, there are way too many cache misses involved
nice
Awesome video
Thanks! Glad you thought so! :^)
What a wonderful video! Simple but great example, though we need some C++20 features such as span, binary_semaphore, jthread.
Btw, I guess -pthread would be better than the -lpthread flag!
Thanks for the great explanation ;)
This guy has a "C++ from scratch" playlist, where he explains all these C++20 features. That playlist is a pre-requisite for these videos.
Good stuff!
Glad you enjoyed it! :^)
Wonderful
Very useful, thanks!
Glad you liked it!