In Engineering, when you are trying to get into a PhD program, how much does a good Master's degree GPA help you if you have a poor undergrad GPA?

[lserkfjlsdkfmweo]

I asked this on Reddit already, but I thought that I might get some more insights here.

I've talked to a few people in biology fields who didn't do well in undergrad but did very well in their MS programs and were able to attend top ranked schools for PhD. Is this similar with engineering?

I am specifically talking about top ranked schools, such as the caliber of MIT and Stanford. I have about a 3.3 undergrad GPA, but have a great research experience with a first author paper on the way and two non-first author papers already published, so I think that I should already be capable of getting into decent schools. However, I want to shoot for the top schools, so I have considered getting an MS degree and trying to get a good GPA there, while continuing research. Is my attempt to make up for it possible, or would the low undergrad GPA screw me forever?

Some people have told me that, due to the higher curve in MS programs, the GPAs obtained from MS programs are not seen as seriously. If this is the case, is another BS degree a better use of my time than an MS degree? Not to mention that a BS would take longer so I would have more time to accomplish things, and a lot of cool honors and awards seem to be only available to BS students and are not for MS students. There's also the fact that more than a certain amount of grad school will disqualify me from applying to NSF.

Edited May 29, 2016 by lserkfjlsdkfmweo