Optimize LQ::probs for large number of targets

[vincentmr]

Before submitting

Please complete the following checklist when submitting a PR:

• All new features must include a unit test.
  If you've fixed a bug or added code that should be tested, add a test to the
  tests directory!

• All new functions and code must be clearly commented and documented.
  If you do make documentation changes, make sure that the docs build and
  render correctly by running make docs.

• Ensure that the test suite passes, by running make test.

• Add a new entry to the .github/CHANGELOG.md file, summarizing the
  change, and including a link back to the PR.

• Ensure that code is properly formatted by running make format.

When all the above are checked, delete everything above the dashed
line and fill in the pull request template.

---

Context:
The parallel performance of probs is terrible when the input has many targets. For instance with 16 OpenMP threads:

tgt=8,qubits=20,probs,t_avg=5.50773,t_ratio=1/1000
tgt=9,qubits=20,probs,t_avg=8.44628,t_ratio=1/1000
tgt=10,qubits=20,probs,t_avg=17.0945,t_ratio=1/1000
tgt=11,qubits=20,probs,t_avg=43.9831,t_ratio=1/1000
tgt=12,qubits=20,probs,t_avg=135.431,t_ratio=1/1000
tgt=13,qubits=20,probs,t_avg=354.502,t_ratio=1/1000
tgt=14,qubits=20,probs,t_avg=1099.4,t_ratio=1/1000
tgt=15,qubits=20,probs,t_avg=3680.11,t_ratio=1/1000
tgt=16,qubits=20,probs,t_avg=10363.3,t_ratio=1/1000
tgt=17,qubits=20,probs,t_avg=35061.4,t_ratio=1/1000

Description of the Change:
Parallelize the double loop over the loop on probs only.

Benefits:
Faster execution for any number of targets on "normal" processor (< 129 cores). I won't show a proper sweep over the number of targets because the current version is too slow to do so.

tgt=8,qubits=20,probs,t_avg=2.93886,t_ratio=1/1000
tgt=9,qubits=20,probs,t_avg=3.17052,t_ratio=1/1000
tgt=10,qubits=20,probs,t_avg=2.86489,t_ratio=1/1000
tgt=11,qubits=20,probs,t_avg=2.61477,t_ratio=1/1000
tgt=12,qubits=20,probs,t_avg=2.40587,t_ratio=1/1000
tgt=13,qubits=20,probs,t_avg=2.23169,t_ratio=1/1000
tgt=14,qubits=20,probs,t_avg=2.08594,t_ratio=1/1000
tgt=15,qubits=20,probs,t_avg=1.97252,t_ratio=1/1000
tgt=16,qubits=20,probs,t_avg=2.30553,t_ratio=1/1000
tgt=17,qubits=20,probs,t_avg=2.44993,t_ratio=1/1000

Possible Drawbacks:
"Abnormal" processors (i.e. 1024 cores) could benefit from a different parallelization strategy in certain edge cases.

Related GitHub Issues:
[sc-69165]

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 92.32%. Comparing base (356bdce) to head (78fc12f).
Report is 91 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #807      +/-   ##
==========================================
- Coverage   94.25%   92.32%   -1.93%     
==========================================
  Files         131       73      -58     
  Lines       19792    11172    -8620     
==========================================
- Hits        18655    10315    -8340     
+ Misses       1137      857     -280     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[maliasadi]

Thanks @vincentmr!

[LuisAlfredoNu]

Good improvement with OPENMP