1. 5

[Summarizing since lobste.rs doesn’t add summaries from YT] Interview with a guy from a startup called GPU Audio that’s doing exactly that: audio processing on GPUs. He claims a 40x performance boost over CPU-based processing. They’ve done a lot of work on parallelizing the algorithms while keeping latency low. This is also very scalable for running large numbers of audio streams, as on a server or implementing modular synthesizers.