Is your feature request related to a problem? Please describe.
currently the premerge tests take ~30min partly because of the tedious building process with options such as
-gencode=arch=compute_52,code=sm_52 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -gencode=arch=compute_80,code=sm_80 -gencode=arch=compute_86,code=compute_86 -gencode=arch=compute_86,code=sm_86

Describe the solution you'd like
quick test should be built just use a specific cuda arch