- Yes, most PSUs can send much more than 150W over the 8-pin cable. Did you know that the extra two pins on 8-pin PEG are for... ground! So a 6-pin connector is rated for 75W, add in two more ground pins and it's double the power delivery. Meaning the 12V lines are easily capable of delivering 300W on most good quality PSUs. But what about low quality PSUs, which absolutely do exist? There will inevitably be some PSUs with a single 8-pin connector on the end of a cable and the cable won't really be rated for more than 150W delivery. If there were two 8-pin connections, there would be less risk.
- There's no such thing as an "8-pin Molex connector" or "molex microfit 12-pin." Molex was a company that designed early interconnect systems for electronics, and the 4-pin Molex connector became the standard for PCs back in the 80s. But no one ever calls the modern 6-pin, 8-pin, and 12-pin connectors "Molex" because the company had nothing to do with designing them. (See, I can also be pedantic.)
- The main point is that it would have been trivial for Nvidia to add dual 8-pin inputs, as it already had the adapter cables available and economies of scale would come into play.
- You're seriously going to say that a graphics card that slots into a PCIe slot and uses PCIe power connectors for auxiliary power doesn't actually need to conform to the PCIe standard?
More critically, it was rather stupid to push an early 12-pin connector, even on cards that don't need it, and then include an adapter in every card. Notice how many of the AIC partners used 12-pin connections for RTX 30-series? Exactly none. Not a single one. There are now, more than 18 months later, RTX 3090 Ti cards that use... a 16-pin connector, which is basically the same thing as Nvidia's 12-pin plus four extra signaling pins. Still, it's an official standard at least. And all of the 3090 and lower GPUs that aren't Founders Edition models will still use dual 8-pin connectors, because it really doesn't simplify card design or improve the cards in any measurable way.
It's like Nvidia crowing about how awesome the new cooling design is for the RTX 30-series (mostly the 3080), but then all of the AIC partners used traditional cooling designs that in most cases cool better than the Founders Edition equivalent. Why? Because Nvidia spent time and money on making a card that looks different, rather than a card that actually performs better. 12-pin was different for the sake of being different, not because it was actually necessary.