So, here's some data:
Code: Select all
P-State Multiplier Freq (GHz) Stock (V) Under (V)
Boost 2 20.0 4.0 1.425 1.325
Boost 1 18.5 3.7 1.400 1.225
0 17.5 3.5 1.325 1.175
1 14.5 2.9 1.225 1.050
2 11.5 2.3 1.125 0.925
3 8.5 1.7 1.000 0.825
4 7.0 1.4 0.8875 0.775
------------------------------------------------------------------------------
The next bit are power numbers. I measured using a Kill-A-Watt at the wall before my UPS. That includes my monitors, computer box, speakers, and such. It's not ideal, but my UPS didn't like providing power if I didn't have the box in the main spot. Deltas are what matter, so that'll do.
Code: Select all
Stock Voltage Undervolted Deltas
Speed (GHz) Threads Idle (W) Load (W) Idle (W) Load (W) Idle (W) Load (W)
1.4 8 200 232 197 225 1 7
1.7 8 205 249 197 234 7 15
2.3 8 212 279 201 254 11 25
2.9 8 220 316 204 283 16 33
3.5 8 235 362 218 320 17 42
3.7 8 250 401 223 335 27 66
Unlimited 2 196 292 196 268 0 24
Unlimited 4 196 336 196 291 0 45
Unlimited 8 196 348 196 306 0 42
Not sure that I'll leave it this way, but it is interesting, to say the least. It is especially interesting to consider how it would affect people's opinions of AMD chips if AMD was more aggressive on voltage and lowered their TDPs.
I did end up having a couple of freezes (not BSODs, just freezes) after I set these to run in normal operation, which occurred in under 15 minutes in every case. This, despite having proven each P-state for over 12 hours on Prime95 (some were well over 12 hours, but 12 was the minimum for some transition states). By upping the NB frequency from the default 2200 MHz to 2400 MHz (as I'd had it before I did any tweaking, for memory optimization), I seem to have resolved that issue.