I'm not arguing that the grounding actually works, but...
Silicon is not an insulator. It is a semiconductor. Plus, at the very fast switching speeds and low voltages, we have measured performance differences at work on whether our chuck tops were voltage biased, connected to ground or left floating. We produce semiconductor testing equipment where you put a fully processed (or not fully processed, as the case might be) silicon wafer onto a metal chuck (flat round surface) for electrical testing.
Now, when I say performance, I don't mean overclocking performance. I'm talking about threshold differences that might mean NOTHING to a CPU chip. These are "performance differences of femto-amps at 110GHz frequency and voltage sweeps. It's the capacitance problem. At high frequencies, capacitance issues degrade the edges of logic signals and filter out high frequency components to analog signals (particularly important to microwave frequency amplifier/processing components).
I'm saying all this because I think the subject is fascinating, not because I think it actually makes a difference. In commercial chip designs, the designers assume a worst case capacitance based upon a normal air-cooled (ungrounded) design. So, there is probably PLENTY of headroom to deal with stuff like that.
Who knows....
|