TDP这个用来描述处理器功耗水平的参数在今天其实已经有点不准了，Intel和AMD两家都允许旗下部分处理器短时间内越过TDP限制进行自动超频，比如Intel从Sandy Bridge时代引入的Turbo Boost 2.0技术就是这样，在引入这项技术的同时，也引入了新的Power Level概念。
一般来说，PL1=TDP，PL2则是在睿频状态下处理器的功耗墙，不能超过它。另外还有PL3和PL4，都是处理器瞬间功耗，这两项一般被锁死，用户是不能手动调节的。除了功耗参数外，还有一个持续时间的参数Tau，它用来规定处理器能在PL2状态下工作的时间。Intel并不会直接对外公布这几项数据，而是将其写在处理器的datasheet里面。在最新一版的Comet Lake处理器datasheet中，Intel公开了Comet Lake-S系列处理器的详细功耗设定数据，数据整理如下：
预览var tableid = settableid();
document.write('');
Table().init({
title: "Comet Lake-S PL1 PL2 Tau规格表",
header: ["分类","TDP","PL1","PL2","Tau"],
width: '40%',
padding: '8px 15px',
columnWidth: 'MAX', //表格单元宽度，"MAX"最大内容显示，""AVG"平均，"AUTO"自动，指定[15,16,20]
align: 'left',
id: tableid,
data: [
["十核带K","125","125","250","56"],
["八核带K","125","125","229","56"],
["六核带K","125","125","182","56"],
["十核","65","65","224","28"],
["八核","65","65","224","28"],
["六核","65","65","134","28"],
["四核","65","65","90","28"],
["双核","58","58","58","28"],
["十核带T","35","35","123","28"],
["八核带T","35","35","123","28"],
["六核带T","35","35","92","28"],
["四核带T","35","35","55","28"],
["双核带T","35","35","42","28"]
]
});核心数量对PL2的默认值有较大的影响，可以看到十核和八核处理器的PL2设定要比其他处理器高上不少，另外PL2/PL1这个系数也放大不少，在八代、九代的Coffee Lake上面，PL2一般都是PL1 x 1.25。带K处理器不仅有更高的PL1，还会有更高的PL2和长达两倍的Tau时间，就算用户不手动设定，它的表现也会比以前的带K处理器好一些。而对于不带K的处理器，PL2/PL1系数最高去到了3.45，十核和八核型号的PL2也去到了224W之高，对众多B460主板来说，是个严峻挑战。
