Simplify library wrappers by introducing a CUDA Ref

[maleadt]

No description provided.

[codecov]

Codecov Report

Merging #414 into master will increase coverage by 0.05%.
The diff coverage is 96.04%.

@@            Coverage Diff             @@
##           master     #414      +/-   ##
==========================================
+ Coverage   78.78%   78.83%   +0.05%     
==========================================
  Files         166      170       +4     
  Lines        8787     8823      +36     
==========================================
+ Hits         6923     6956      +33     
- Misses       1864     1867       +3     

Impacted Files	Coverage Δ
lib/cudnn/conv.jl	68.29% <0.00%> (ø)
lib/cusparse/level2.jl	84.52% <ø> (ø)
lib/cusparse/level3.jl	77.31% <ø> (ø)
test/cublas.jl	91.66% <ø> (ø)
test/runtests.jl	50.00% <ø> (ø)
lib/cudnn/batchnorm.jl	38.29% <66.66%> (ø)
src/pointer.jl	75.38% <75.00%> (-2.67%)	⬇️
lib/cublas/wrappers.jl	91.28% <86.66%> (-0.65%)	⬇️
test/cutensor/elementwise_trinary.jl	96.92% <96.92%> (ø)
lib/cublas/CUBLAS.jl	82.97% <100.00%> (ø)
... and 23 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 e9beb5e...7b62c35. Read the comment docs.