AVX2 fused SqrClippedReLU + ClippedReLU
Speedup measured with profile-build x86-64-avx2:
```
Result of 50 runs
==================
base (./sf-master ) = 777661 +/- 8102
test (./sf-fused ) = 796739 +/- 9127
diff = +19077 +/- 3164
speedup = +0.0245
P(speedup > 0) = 1.0000
CPU: 8 x AMD EPYC 7R13 Processor
```
Passed STC:
https://tests.stockfishchess.org/tests/view/6a5049335529b8472df8013c
LLR: 2.96 (-2.94,2.94) <0.00,2.00>
Total: 262624 W: 68444 L: 67815 D: 126365
Ptnml(0-2): 724, 28976, 71325, 29521, 766
closes https://github.com/official-stockfish/Stockfish/pull/6976
No functional change