My question is just as the title. Actually, I am looking for a way to free device memory asynchronously.
Thanks!
cudaFree()
is synchronous. If you really want it to be asynchronous, you can create your own CPU thread, give it a worker queue, and register cudaFree
requests from your primary thread.
That said, asynchronous frees seem like an odd request. Perhaps you could explain why you'd like it to be asynchronous. Do you want the free to occur immediately after a CUDA event is triggered?
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With