As someone who bought a 2070 9 months ago for 500$, I have nothing to complain about. This is technology progress, process maturity and competition. NVidia is acting as any other corporation would, including AMD.
I have been saying it for a while. NVidia's RTX 20 series prices cannot be judged without taking the silicon size, and accordingly the yields factor, into consideration. An RTX 2070 or 2070 Super with die sizes north of 450mm2 at 500$ versus AMD Navi on 251mm2 at , an 450$, and it's almost guaranteed AMD has the better profit margin.
Now bear in mind NVidia always had the upper hand when it came to extracting better gaming performance from the same die size, also a factor in its superior power efficiency. This generation, they decided to bet on RT/Tensor cores, and whatever your take on the implementation is, it remains the right bet imho. Hybrid rendering technique was approaching a level of maturity where it can be "doable" in triple AAA titles. Their cards were getting really fast enough that an alternative GTX 2080 ti with a 750mm2 die area entirely dedicated to raster would be approaching the point of an overkill for most of games even at 4K resolution, which meant they needed another technology to drive them forward. On the other hand, AMD could have jeopardized their plans if they had a competing product comparable in efficiency and output, available at the right time with the right quantities. But they didn't, which meant NVidia could go ahead with their plans.
It was the perfect plan, at the perfect time, and it can only get better with the next gen 7nm cards.
I have been saying it for a while. NVidia's RTX 20 series prices cannot be judged without taking the silicon size, and accordingly the yields factor, into consideration. An RTX 2070 or 2070 Super with die sizes north of 450mm2 at 500$ versus AMD Navi on 251mm2 at , an 450$, and it's almost guaranteed AMD has the better profit margin.
Now bear in mind NVidia always had the upper hand when it came to extracting better gaming performance from the same die size, also a factor in its superior power efficiency. This generation, they decided to bet on RT/Tensor cores, and whatever your take on the implementation is, it remains the right bet imho. Hybrid rendering technique was approaching a level of maturity where it can be "doable" in triple AAA titles. Their cards were getting really fast enough that an alternative GTX 2080 ti with a 750mm2 die area entirely dedicated to raster would be approaching the point of an overkill for most of games even at 4K resolution, which meant they needed another technology to drive them forward. On the other hand, AMD could have jeopardized their plans if they had a competing product comparable in efficiency and output, available at the right time with the right quantities. But they didn't, which meant NVidia could go ahead with their plans.
It was the perfect plan, at the perfect time, and it can only get better with the next gen 7nm cards.