In short...pic.twitter.com/KPLuYBr1cz
You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
If It was that simple, $AMD would not have spent so much marketing dollars to put their own definition of TDP ... It is extremely confusing to consumers.
If you want to understand why TDP is not the right matrix, here it is … I was one of those giving this lecture for 1 year, so, here it is again (I posted it 2 days ago already) https://www.intel.com/content/dam/doc/white-paper/resources-xeon-measuring-processor-power-paper.pdf … It is a lot more complicated than it looks.
If you are interested on the modern way to figure out your power models, there was SDP too, it got kind of erased from http://intel.com , TDP and ADP, SDP can all be manipulated, based on the scenario you want to benchmark https://www.intel.com/content/dam/www/public/us/en/documents/design-guides/atom-e3800-m-d-i-soc-thermal-design-guide.pdf …
For chip designers, TDP is just a average number, you can actually use a lot more than the TDP, most laptop CPU do that, and Servers chips do too, most people benchmarking servers don't really test those either, as they trend to "average their results to express opinions"
But for a matter of fact, a processor like the Xeon Gold, equipped with Hardware P state (Speedshift) will answer a database request a lot faster than a CPU without it, while using more power than the TDP for few milliseconds … You see … it is not only about reading only ...
The Devil is in the details, with hardware P-state, the speed for database request is a lot better, most #responsible system designer will not run its CPU at 100% load, they run them at 70% … giving room for #speedshift to do its work, because nobody design systems to max out
But the press keep testing their system at 100% load, neglecting the entire spectrum of usage models that are the majority of system out there. Very few people design their server farms to have a node max out at day one. (if you do, you are not too smart)
Bottom line, The scenario for your new server system does include leaving room for expension of your compute capability, then, SDP is definitively more important than TDP, and performance tricks like Hardware P State are the real boosters of your user experience latency.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.