That's okay, this is fun for me too. ANd thanks, I thought something was off.I think you forgot the "^3" on the last one. Not really wanting to hold on to this, but I've seen how different silicon behaves under different situations and they certainly do not scale identically. The results can be all over the place.
4.725*4 + 101.89 = 120.89W
vs 96W on Anandtech
So that tells me perhaps it's a little higher.
If I go with my initial assumption of Po of 30W.
30W + (39*2.033) = 109W vs 96W
I also need to account for DVFS, which will make things harder to figure out.
It does appear at the levels of
3600 95
3700 96
3800 96
There is something kicking in that is keeping power constant. That pattern goes away at 3900 and continues to be exponential after that.
Since we're on the topic; this is cool!
Sort of a neat snap shot of the way things have been moving. That frequency capped off quick, and we've been trending slightly downwards while cores continue on a linear path upwards, but tapering off as we hit size limits.
There is a way forward, it seems like AMD is banking on chiplets. Curious to see what other solutions will arrive.
Last edited: