Add Support for Heterogeneous GPU Configurations in the Cuda Component

[Treece-Burgess]

Pull Request Description

This PR adds support for heterogeneous gpu configurations. As a consequence the following were also updated:

• How we internally handle the default device id
• How we handle creating a context for users if one is not provided

Tested on Leconte (8 * V100) and Hexane (1 * V100 & 1 * H100):

Test	Pass
HelloWorld.cu	✅
HelloWorld_noCuCtx.cu	✅
concurrent_profiling.cu	✅
concurrent_profiling_noCuCtx.cu	✅
cudaOpenMP.cu	✅
cudaOpenMP_noCuCtx.cu	✅
pthreads.cu	✅
pthreads_noCuCtx.cu	✅
simpleMultiGPU.cu	✅
simpleMultiGPU_noCuCtx.cu	✅
test_2thr_1gpu_not_allowed.cu	✅
test_multi_read_and_reset.cu	✅
test_multipass_event_fail.cu	✅

Author Checklist

• Description
  Why this PR exists. Reference all relevant information, including background, issues, test failures, etc
• Commits
  Commits are self contained and only do one thing
  Commits have a header of the form: module: short description
  Commits have a body (whenever relevant) containing a detailed description of the addressed problem and its solution
• Tests
  The PR needs to pass all the tests

[dbarry9]

I have tested this PR on two configurations:

• V100+A100
• V100+H100

I monitored accurate counts of FP32 and FP64 events on both systems.