[Performance] Do not fuse neighbor sampler for 1 thread

[anko-intel]

Description

It fixes the issue #6315 (at least for CPU-only configuration)

The fused version of NeighborSampler (sample_neighbors_fused) is prepared as multithreaded, so we can not use it when the number of available threads is lower than 2.
This way we can achieve the same performance as previously with not fused version (which is also multithreaded but with less overhead).

Performance results of the script from the issue #6315 (modified to be allowed to run on CPU-only configuration) on AWS EC2 instance r6i.32xlarge (2 socket Intel(R) Xeon(R) Platinum 8375C) are as follow :

Checklist

Please feel free to remove inapplicable items for your PR.

• The PR title starts with [$CATEGORY] (such as [NN], [Model], [Doc], [Feature]])
• I've leverage the tools to beautify the python and c++ code.
• The PR is complete and small, read the Google eng practice (CL equals to PR) to understand more about small PR. In DGL, we consider PRs with less than 200 lines of core code change are small (example, test and documentation could be exempted).
• All changes have test coverage
• Code is well-documented
• To the best of my knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change
• Related issue is referred in this PR
• If the PR is for a new model/paper, I've updated the example index here.

[dgl-bot]

To trigger regression tests:

• @dgl-bot run [instance-type] [which tests] [compare-with-branch];
  For example: @dgl-bot run g4dn.4xlarge all dmlc/master or @dgl-bot run c5.9xlarge kernel,api dmlc/master

[dgl-bot]

Commit ID: 05613e1

Build ID: 1

Status: ✅ CI test succeeded.

Report path: link

Full logs path: link