[CUDA] Upgrade to PTX 6.3 and add a few CUDA intrinsics

[yuanming-hu]

[Click here for the format server]

---

[codecov]

Codecov Report

Merging #1548 into master will increase coverage by 0.11%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #1548      +/-   ##
==========================================
+ Coverage   67.33%   67.44%   +0.11%     
==========================================
  Files          38       40       +2     
  Lines        5596     5625      +29     
  Branches      978      981       +3     
==========================================
+ Hits         3768     3794      +26     
- Misses       1660     1662       +2     
- Partials      168      169       +1     

Impacted Files	Coverage Δ
python/taichi/tools/file.py	0.00% <0.00%> (-28.58%)	⬇️
python/taichi/__init__.py	75.00% <0.00%> (-17.00%)	⬇️
python/taichi/lang/impl.py	89.41% <0.00%> (-1.20%)	⬇️
python/taichi/misc/util.py	44.44% <0.00%> (-0.15%)	⬇️
python/taichi/core/__init__.py	100.00% <0.00%> (ø)
python/taichi/misc/settings.py
python/taichi/misc/np2ply.py
python/taichi/core/settings.py	48.14% <0.00%> (ø)
python/taichi/tools/__init__.py	100.00% <0.00%> (ø)
python/taichi/misc/__init__.py	100.00% <0.00%> (ø)
... and 7 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update cc621a0...86eef06. Read the comment docs.