Clemens Lutz

Another use-case (that I'm currently facing) is measuring hardware performance counters and GPU event timers. It would be really helpful if Criterion could provide support for user-specified timers/counters. Would it...

This issue also affects PowerPC64, see https://github.com/bheisler/RustaCUDA/pull/54.

I'm successfully using `load_from_file` to load a fatbin. You can generate a fatbin by passing '-fatbin' to nvcc. How are you compiling the binary?

`Drop` fails frequently during development. If you do illegal memory accesses inside your GPU kernel, you will get something like this: ``` thread 'main' panicked at 'Failed to deallocate CUDA...

> Previously, I thought `AsyncCopyDestination` might be able to take `T: DeviceCopy` references as values for `copy_from` and `copy_to`, just like `CopyDestination`. However, it appears not using page-locked memory will...

Following up on @gaohao95's comment, our use case is storing TPC-H data in a `cudf::table`. At scale factor 100, the `l_comment` and `ps_comment` string columns overflow the 32-bit offset.