Sandra is a miserable hard drive benchmark, and drives are extremely hard to benchmark with any degree of accuracy and repeatability across different machines even when the test is not so clearly flawed. This simply doesn't mean a thing.
That's not really a hard drive test at all. It's interesting that you get different results, and it may point to slight buffer differences. But if anything it just serves to point out how little interface differences amount to in real terms, however much wind they may blow up a particular benchmark's skirt. But this function involves the hard drive to such a small degree and is showing a difference that is likely within the margin of error of the test, and as such its hard to draw many conclusions from it.