- Joined
- Sep 5, 2023
- Messages
- 334 (0.81/day)
- Location
- USA
System Name | Dark Palimpsest |
---|---|
Processor | Intel i9 13900k with Optimus Foundation Block |
Motherboard | EVGA z690 Classified |
Cooling | MO-RA3 420mm Custom Loop |
Memory | G.Skill 6000CL30, 64GB |
Video Card(s) | Nvidia 4090 FE with Heatkiller Block |
Storage | 3 NVMe SSDs, 2TB-each, plus a SATA SSD |
Display(s) | Gigabyte FO32U2P (32" QD-OLED) , Asus ProArt PA248QV (24") |
Case | Be quiet! Dark Base Pro 900 |
Audio Device(s) | Logitech G Pro X |
Power Supply | Be quiet! Straight Power 12 1200W |
Mouse | Logitech G502 X |
Keyboard | GMMK Pro + Numpad |
I'm on an EVGA z690 Classified that hasn't had a BIOS update in quite some time (thanks for riding into the sunset, EVGA /s). I likely won't get the microcode update. Also, I've been running a launch-day 13900k at the same settings for a couple years now and so far *knocks on wood*, it's been fine. I can't set max VID on this board without doing an override voltage that would force higher than normal all-core voltage, but not enough voltage for boosting.
My settings, for reference, but they're probably not helpful to anybody with another OEM's motherboard:
I have it on adaptive "auto", but I tweaked the V/F points at 5100, 5400, and 5700 to -40. If I leave the BIOS at default, it'll boost voltages higher, but for some reason this setting seems to work. Their latest BIOS release sort of...barely...has functioning V/F point control, so shifting these three points down dropped the whole curve a bit. Their BIOS doesn't tell you the actual voltage points, it just lets you put in offsets. Also, if you put in too big of a global offset it just gets ignored, but manually setting "adaptive" and "auto" on that actually drops the voltage from leaving voltage mode as "auto" instead of "adaptive" (*shrugs*). I also have it set up to boost to 6GHz on 1-4 active cores, then 5-8 active are set to 5.8, 5.7, 5.6, 5.5GHz. e-Cores are at 4.6GHz. Ring 4.6GHz. FLL Override Mode 1. Vdroop on -75% (min droop). PWM on 600. Admittedly I got the FLL, Vdroop, and PWM settings from Loomi's suggestions on Youtube for these series of boards and all I've done to verify them is try a few other settings and confirm that these are more stable.
So the end result:
I don't have a good screenshot of HWiNFO64, but VIDs end up in the 1.25-1.325ish range for most tasks, especially large workloads that use lots of cores. If i run lots of low-thread workloads I'll see it peak around ~1.4-1.417V at 6GHz.
Also, there was an older microcode that ran lower voltages, and thus temps were lower...I'm currently thinking about doing some testing to see if I can run that one. On that one I just had the V/F points mentioned above at -25. The BIOS update thread on their forums indicated some stability improvements with the newer microcode so I've just been using it (for...over a year?...I don't remember when that version came out). I did see that thread referenced in GN's video actually, but they seemed to have the information backwards because the newer microcode increased the voltage (and heat).
Benchmate capture from pushing it a little (I think I may have had 5.6 or 5.7GHz at 8-cores, other settings above same) shows 1.37V on P-cores, but I don't know if that's max-VID:
Really pushing it a while back I did this CB23 run, with P-core voltage showing 1.42V.
For most gaming workloads though, just on my 24/7 settings mentioned earlier, it usually runs around 115-140W and it mostly stays below 1.4V (that being the max number in HWiNFO).
My settings, for reference, but they're probably not helpful to anybody with another OEM's motherboard:
I have it on adaptive "auto", but I tweaked the V/F points at 5100, 5400, and 5700 to -40. If I leave the BIOS at default, it'll boost voltages higher, but for some reason this setting seems to work. Their latest BIOS release sort of...barely...has functioning V/F point control, so shifting these three points down dropped the whole curve a bit. Their BIOS doesn't tell you the actual voltage points, it just lets you put in offsets. Also, if you put in too big of a global offset it just gets ignored, but manually setting "adaptive" and "auto" on that actually drops the voltage from leaving voltage mode as "auto" instead of "adaptive" (*shrugs*). I also have it set up to boost to 6GHz on 1-4 active cores, then 5-8 active are set to 5.8, 5.7, 5.6, 5.5GHz. e-Cores are at 4.6GHz. Ring 4.6GHz. FLL Override Mode 1. Vdroop on -75% (min droop). PWM on 600. Admittedly I got the FLL, Vdroop, and PWM settings from Loomi's suggestions on Youtube for these series of boards and all I've done to verify them is try a few other settings and confirm that these are more stable.
So the end result:
I don't have a good screenshot of HWiNFO64, but VIDs end up in the 1.25-1.325ish range for most tasks, especially large workloads that use lots of cores. If i run lots of low-thread workloads I'll see it peak around ~1.4-1.417V at 6GHz.
Also, there was an older microcode that ran lower voltages, and thus temps were lower...I'm currently thinking about doing some testing to see if I can run that one. On that one I just had the V/F points mentioned above at -25. The BIOS update thread on their forums indicated some stability improvements with the newer microcode so I've just been using it (for...over a year?...I don't remember when that version came out). I did see that thread referenced in GN's video actually, but they seemed to have the information backwards because the newer microcode increased the voltage (and heat).
Benchmate capture from pushing it a little (I think I may have had 5.6 or 5.7GHz at 8-cores, other settings above same) shows 1.37V on P-cores, but I don't know if that's max-VID:
Really pushing it a while back I did this CB23 run, with P-core voltage showing 1.42V.
For most gaming workloads though, just on my 24/7 settings mentioned earlier, it usually runs around 115-140W and it mostly stays below 1.4V (that being the max number in HWiNFO).