Honestly, I think what annoys/concerns a lot of people is that the peak consumption numbers are considerably higher (~90W and ~115W, respectively), and that the BIOS on a lot of motherboards does not respect Intel's maximum turbo duration by default (because it's effectively "free performance"), which leads to noticeably higher-than-expected power consumption on long-running high-utilisation workloads. Obviously, things are more complex when you start to consider total power consumption to complete task, etc..
Ryzen also has a higher PPT, which roughly corresponds to PL2, as outlined above, though. The factor is usually 1.35, landing a 65W Ryzen at around 88W, and I heard of people who tuned their chips to the point where they got prettyuch into Intel levels of power draw. It's not like that is only an Intel thing, yet Intel is the one being critized pretty much exclusively for it. Yes, it would be better if they actually stuck to the limits, especially considering that the performance drop isn't that huge while drastically increasing efficiency. There were tests that showed that, too. And Raphael will play pretty much in Intel's ballpark, too, but few people actually seem to mind while Intel is mocked for it all over the place.
Another great way to reign in Alder Lake power consumption is undervolting, though. My 12700k draws 135W at max in Cinebench 23 and the stock 4.7GHz all-core turbo clock, due to my undervolt. I also ran it with a slight OC of 100MHz for a time just to test it. It gave me less than 2% more performance, for 30W more power consumption, so I scrapped it again. 135W is well within the PPT power limits of a Ryzen 9 5900X CPU, while also beating it in that benchmark. It's also only 10W above PL1, and it stands to reason that setting power limits appropriately would result in a very minor performance loss again. Now, to be fair, I realize that my chip might be simply a pretty well-binned one; I actually had that suspicion for a while now. My undervolt is an adaptive plus offset set to 1.24V and -0.135V, respectively; it might be 1.28V, though, I'm not 100% sure and can'tlook right now, but it's somewhere in that ball park. I have seen people on Youtube with worse results, implying that either their chips are not undervolt happy, or mine is exceptionally so. Other benchmarks continuously throw out scores on the higher end of results for my chip, so I'm positive there that it's good silicon.
Facebook
Twitter
Instagram