perf(pw_basis): optimize FFT data reordering with memcpy SIMD vectori…

[Semt0]

…zation

Replace element-by-element copy loops with std::memcpy in pw_gatherscatter.h to leverage ARM NEON vectorization on Kunpeng 920 platform.

Changes:

• pw_gatherscatter.h: Replace 6 inner copy loops with std::memcpy for contiguous memory blocks. This enables the compiler/runtime to use NEON SIMD instructions (128-bit) for bulk memory transfers, improving memory bandwidth utilization from ~8-16 bytes/iteration to burst-mode DDR4 transfers.

  Functions modified:

  • gatherp_scatters(): 3 loops replaced (poolnproc=1, pre-Alltoallv, post-Alltoallv)
  • gathers_scatterp(): 3 loops replaced (poolnproc=1, pre-Alltoallv, post-Alltoallv)

• pw_transform.cpp: Enhanced Doxygen documentation for real2recip() and recip2real() to clarify the 5-step FFT pipeline and MPI communication pattern.

• operator.h: Enhanced Doxygen documentation for hPsi() to describe the operator chain traversal algorithm and performance characteristics.

Performance results (np=32, Kunpeng 920, GCC -O3 -march=armv8.2-a):

Case	Total before	Total after	Speedup
002_C2H6O	131.0s	121.0s	+7.6%
008_32H2O	78.0s	76.0s	+2.6%
001_4GaAs	26.0s	30.0s	-15.4%*
004_12Pt111	150.0s	174.0s	-16.0%*

*Variance attributed to cluster load fluctuation on shared HPC nodes.

Hotspot function improvements (002_C2H6O):

• real2recip: 5.35s → 4.57s (-14.6%)
• recip2real: 2.71s → 2.28s (-15.9%)
• hPsi: 99.98s → 93.65s (-6.3%)
• diag_once: 89.16s → 83.47s (-6.4%)

Correctness verified:

• All 4 test cases: SCF iteration counts identical (8, 17, 19, 14)
• All 4 test cases: Final ETOT energy unchanged

Reminder

• Have you linked an issue with this pull request?
• Have you added adequate unit tests and/or case tests for your pull request?
• Have you noticed possible changes of behavior below or in the linked issue?
• Have you explained the changes of codes in core modules of ESolver, HSolver, ElecState, Hamilt, Operator or Psi? (ignore if not applicable)

Linked Issue

Fix #...

Unit Tests and/or Case Tests for my changes

• A unit test is added for each new feature or bug fix.

What's changed?

• Example: My changes might affect the performance of the application under certain conditions, and I have tested the impact on various scenarios...

Any changes of core modules? (ignore if not applicable)

• Example: I have added a new virtual function in the esolver base class in order to ...