• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Ryzen Master Performance Boost Overdrive EDC vs. PPT showdown

Joined
Jun 3, 2010
Messages
784 (0.22/day)
Hi,
https://www.reddit.com/r/Amd/comments/cnykxh/3700x_frequencyvoltajepower_on_pbo_custom/

I just found this on Reddit which I happened to consider to be useful for comparison's sake between EDC vs. PPT settings. Normally, you would think power regulation via PPT should fare better. That is more than likely, not to be the case. PPT helps relieve socket TDP for untampered overclocking. EDC is better in undervolting. I hope that helps.
Here are the charts reordered for better visibility of real world performance on Cinebench-R15. Not to spur any fanbase controversy - it is just that, the evidence base is relatively limited on this field. :p
EDC 3.png
PPT 3.png
 
Joined
Jun 3, 2010
Messages
784 (0.22/day)
Folks, I need help. Marketting accomplice won't help me with this, I need somebody at AMD to see this.
The basic idea is very simple, but effective: EDC ≤ PPT ≤ TDP*. At low EDC threshold, EDC~TDC. A DOCP algorithm can first check effective cooling to type in these various settings automatically. There is a finite amount of system stability amounting to vdroop & temperature, so single measure solution that takes care of the consumer side of the variables could customise deployed performance base.
*: Tdp, that is right, I didn't mistype.
 
Joined
Jun 3, 2010
Messages
784 (0.22/day)
As a conjunct, found this:
https://www.reddit.com/r/Amd/comments/cks7io Comparison;
  1. Optimal point of EDC for stock 3700X(65w) is 80A whereas,
  2. Optimal point of EDC for stock 3900X(95w) is 75A.
This goes to show, cpus are mainly limited by the temperature gradient which itself is tempered by the EDC gradient, imho.
 
Top