Originally Posted by WannaBeOCer
Most of the PC builders who actually pay attention to clock speed have boards that enable MCE by default. Which means Intel users won't see the default base clock when under stress. I have my chip at stock and it runs at 4.2Ghz when under load not 3.6Ghz and I see 5Ghz boost in games often. AMD screwed up their Turbo and they fixed it.
Yeah well new AMD CPUs don't leave over 1GHz left untapped like not so old Intel CPUs. Even new Intel CPUs are not pushed to limit that well out of the box. Intel still has a way to improve their products. Where as AMD is using boost similar to what people are now more used to on modern GPUs, opportunistic boost based on various metrics, not a simple predefined static boost table like Intel. The old approach that left performance untouched was "good" for overclocking and made people "happy" they had something to play with and exploit, where as these newer GPUs/CPUs with much better boost algorithms leave close to nothing to exploit by OC ==> people are not as happy.
I've yet to see Intel platform motherboard with PL1/PL2/... set to it's proper TDP, say 95W instead of 4096W and 1024A current limit. Some did ship with MCE disabled, some. A 9900K run to absolute specs of it's TDP, no MCE, is not that fast and will throttle, turning into an overpriced 8700K like performance. So... reviewers don't care about specs much, slap it on, unlimit power, thermals, etc. and let it rip, often even with MCE on especially when their board enabled MCE on default, or their board pushed 101MHz BCLK and other sneaky OC.
The Turbo on AMD is not really screwed up, it's just impractical to monitor for 99.99% of people accurately for any verification of does it boost to max specs or not. They can tweak the algorithms and try please people with slightly higher boost here and there, and they have done that even though they didn't have to.