Workaround for assignment-after-reduction offload bug

[markdewing]

Loop index over nw must be 32 bits in size, otherwise the assignment following the reduction will miss the second item.
Bug affects offload with NVidia.
See #4767

This PR represents the minimal changes needed to fix the real and complex offload builds for the multideterminant tests.
More changes could be made to make the type of nw consistent (A static analyzer would complain if run against the current code).

What type(s) of changes does this code introduce?

Delete the items that do not apply

• Bugfix

Does this introduce a breaking change?

• No

What systems has this change been tested on?

local server (A2000), JLSE A40

Checklist

Update the following with a yes where the items apply. If you're unsure about any of them, don't hesitate to ask. This is
simply a reminder of what we are going to look for before merging your code.

• Yes. This PR is up to date with current the current state of 'develop'
• Yes. Code added or changed in the PR has been clang-formatted
• No. This PR adds tests to cover any new code, or to catch a bug that is being fixed
• N/A. Documentation has been added (if appropriate)

[ye-luo]

Good fix. We probably better to change all the GPU offload related loop to uint32_t. I think all the GPUs naturally handle 32 bit integer with 64 bit addressing.

[markdewing]

Test this please