Quote Originally Posted by OhNoes! View Post
Thanks, see below.

I appreciate your explanation.

I am aware that a leaky chip would require more voltage to operate at the same clock or frequency, but that is not what my argument is. My argument is about efficiency at utilizing a given power (let's say set in bios) and the resultant heat.

Take 1 board, put a highly leaky chip in there, set voltage to 1.20v. Stress it. Take the chip out, put another less leaky chip in there, repeat the process by setting voltage to 1.20v and stressing that chip also.

My argument is that because of the energy that is being lost through leakage, the leaky, less efficient chip ends up using less power = less heat. The energy that is lost through leakage usually accounts for less heat than if it was actually utilized in the case of the less leaky chip.

In the real world, this translates into less leaky and highly efficient chips with low vids, but hotter temps, and high vid to highly leaky chips that need more voltage same clocks but run cooler at the same voltage. I may be wrong, but in my experience this has always been the norm.
Efficiency and heat output are not really comparable. And when you say energy lost through leakage, there is no such thing as lost energy, energy doesnt disappear, it changes state. Current leakage turns into either 3 things, heat, light or sound. So you have your real world stuff backwards. A higher leakage chip is less efficient due to the fact that it uses more current to operate at the same cpu vcore voltage, where a low leakage chip is more efficient in the use of current and therefore has less excess heat created.

Think of a trough filled 3/4ths of the way with water, and this will be your low leakage chip. The small plastic boat will be the current. Now when you move the water without touching the boat to get it from one side to the other, you dont spill much water out of the trough, but likely a little. Do the same thing to a high leakage chip with the trough this time full to the brim. The same force used to move the boat makes alot of water spill out. Now each chip no mater what produces some heat, and lets call that a constant to make things simple, 48 ounces. Now all the water on the floor next to the tough is the leakage. The more leakage on the floor, the more heat that chip is going to produce as that water didnt help move the little boats to their destination. Now the low leakage chip doesnt have a lot of water on the floor, lets call it 12 ounces, so the constant heat the chip makes plus the leakage equals 60 ounces. Now the high leakage chip has 36 ounces on the floor, plus the constant 48 ounces is 84 ounces. The more ounces, means more heat.

You got to remember that voltage is only one part of the power equation. What causes leakage is the current that is produced. How do you think a toaster works? It runs on 120volts, yet it draws like 15 amps to produce some near 1000watts of power. You see the wires in the toaster glow red and get damn hot. You run lots of equipment at 120volts, but most run at less amperage. The less amps, the less heat, and the less watts. A 60 watt lightbulb runs at 120volts, but it only draws 1/2 an amp. Compare any light bulb, the lower the watts, the less amps they draw all at the same voltage, 120v. The more amps a lightbulb draws, the brighter the light and the more heat the bulb produces. Just as with light bulbs at a certain point it is hard to determine the diff between light produced, but you can damn sure feel the heat difference.

I dont know how i could be more clearer here. Your argument defys the principles of Physics. High leakage chips leak energy, and that leaked energy is heat. The more energy leaked the hotter that chip will be period.