Is Dynamic Parallelism still not supported in latest release?
If not, do we know when we can expect it?
Assuming you’re referring to CUDA Dynamic Parallelism - right now we don’t have the resources to work on this or plan to implement it - do you have a compelling use case that would suggest we think about prioritising it?
Thank you for following up. Well, CUDA’s dynamic parallelism is a holly grail for writing truly expressive & efficient code.