Cupy implementations of dpbench workloads

[adarshyoga]

• Have you provided a meaningful PR description?

This PR adds cupy implementations of six workloads - blackscholes, gpairs, rambo, pca, l2_norm and pairwise_distance. It also contains the necessary changes to the infra to run these workloads. This PR does not contain changes to the CI to run these workloads regularly. For evidence of successful execution, see below report.

• Have you added a test, reproducer or referred to an issue with a reproducer?
• Have you tested your changes locally for CPU and GPU devices?

Tested on orsatdevnuc01 development machine. Below is the test report.

---

Report for 2023-11-02 00:05:57 run

Legend

postfix description device
0 cupy cupy 12th Gen Intel(R) Core(TM) i9-12900
1 dpnp dpnp 12th Gen Intel(R) Core(TM) i9-12900
2 numpy NumPy 12th Gen Intel(R) Core(TM) i9-12900

Summary of current implementation

input_size benchmark problem_preset cupy dpnp numpy
0 20MB black_scholes S Success Success Success
1 8KB gpairs S Success Success Success
2 1MB l2_norm S Success Success Success
3 8MB pairwise_distance S Success Success Success
4 1MB pca S Success Success Success
5 7MB rambo S Success Success Success

Summary of current implementation

input_size benchmark problem_preset cupy dpnp numpy
0 20MB black_scholes S 0.16ms 11.87ms 24.01ms
1 8KB gpairs S 1.66ms 16.81ms 0.67ms
2 1MB l2_norm S 0.08ms 3.75ms 0.37ms
3 8MB pairwise_distance S 0.14ms 2.29ms 3.28ms
4 1MB pca S 11.75ms 10.93ms 2.81ms
5 7MB rambo S 0.49ms 4.53ms 1.27ms

---

• Have you made sure that new changes do not introduce compiler warnings?
• If this PR is a work in progress, are you filing the PR as a draft?

[ZzEeKkAa]

LGTM!