Stop and Frisk was certainly unconstitutional, as applied to anybody of any race, because it was done without any reasonable articulable suspicion that a crime had been committed, was being committed, was about to be committed, or that the "suspect" was armed and dangerous.
But in light of the unconstitutionality of the program, as a practical matter of mathematics/statistics/numbers, if you were going to run an unconstitutional stop-and-frisk program in NYC, with the idea being to totally forget about the requirement for reasonable articulable suspicion [this, rather than "probable cause" is the standard to conduct a Terry Stop] and simply shake down random people to try to turn up guns as a way to reduce gang related shootings, it would have been justifiable, mathematically/statistically, to completely ignore white people because white people were only committing 2% of the gang-related shootings in NYC despite being the largest population group [40%+] in NYC.
With that said, the fact whites were targeted for about 9-10% of searches, despite the fact that they only accounted for 2% of gang-related shootings, that is to say they were being targeted at 5x the rate they were committing the gang-related shooting, suggests that the program, as unconstitutional as it was, lacking any reasonable articulable suspicion, was not a mere "let's shake down dangerous brown people and see what we can find" program motivated by racism.