Implement CUDA backend for parallel `cuda::std::reduce`

[miscco]

This implements a cuda backend for cuda::std::reduce

[github-actions[bot]]

😬 CI Workflow Results

🟥 Finished in 4h 58m: Pass: 82%/134 | Total: 5d 11h | Max: 4h 57m | Hits: 81%/239153

See results here.

[github-actions[bot]]

😬 CI Workflow Results

🟥 Finished in 5h 25m: Pass: 96%/209 | Total: 8d 01h | Max: 5h 23m | Hits: 52%/268885

See results here.

[github-actions[bot]]

😬 CI Workflow Results

🟥 Finished in 3h 37m: Pass: 92%/90 | Total: 3d 05h | Max: 3h 02m | Hits: 41%/194125

See results here.

[github-actions[bot]]

😬 CI Workflow Results

🟥 Finished in 8h 49m: Pass: 95%/134 | Total: 8d 00h | Max: 6h 00m | Hits: 51%/258943

See results here.

[github-actions[bot]]

😬 CI Workflow Results

🟥 Finished in 7h 25m: Pass: 85%/134 | Total: 2d 14h | Max: 6h 00m | Hits: 99%/207194

See results here.

[github-actions[bot]]

😬 CI Workflow Results

🟥 Finished in 2h 54m: Pass: 96%/88 | Total: 3d 02h | Max: 2h 53m | Hits: 31%/201171

See results here.

[github-actions[bot]]

🥳 CI Workflow Results

🟩 Finished in 3h 23m: Pass: 100%/88 | Total: 1d 00h | Max: 1h 38m | Hits: 99%/214282

See results here.