What sort of clocks are you guys getting out of your vRAM specifically?  Like what offset are you finding is stable?  Right now I'm testing at +195/+1100, and so far O_o it's no errors & score is still increasing.

So far I'm at +130/+1000. Everything is stable and I just haven't taken the time to push further. Does the voltage slider do anything? or should I be maxing it out anyways?

arestavo

USMC1419
My PR scores were much higher on XC3 BIOS as well but I went back to stock for now.
 
How do you know how much power the card is drawing on XC3 with the power monitors broken? 
 
I really wonder if it is safe to run this XC3 BIOS on the FTW3 Hybrid long term. I'm high on the list for a KP for the second list, I'm wondering if I should just trade up to that or if it would be worth it with the XC3 BIOS. Not a big benchmarker, just want max FPS in games.

How do I know? I've got an UPS that tells me the power being drawn from the wall, and a titanium PSU for the conversion %. Just gotta do a little math to get a ballpark number.

Is it safe? Don't see why not. It's not an unlocked VBIOS, just a VBIOS that gets the PCIE power draw down into the 60W range and therefor doesn't power throttle like mad. All the HDMI and DP ports work fine as well, unlike other 2-pin VBIOSs.

If you aren't a big benchmarker (I read that as overclocker too), getting a kingpin would be a waste of money. You'd be better off getting an actual XC3 because the performance won't be much different from the stock numbers of both that I've seen. If you do plan on overclocking, the Kingpin will be superior since it had better cooling and a voltage tool not compatible with other cards to gain the stability needed for gaming when heavily overclocking.

Lower pcie power draw is really good to hear! Do you know approximately how much each 8 pin is pulling? Just worried one might be pulling way too much while the another could be pulling too little

glocked89
Lower pcie power draw is really good to hear! Do you know approximately how much each 8 pin is pulling? Just worried one might be pulling way too much while the another could be pulling too little

I don't have that type of equipment to check it. One reads 70 is watts, one 150, the other one is missing completely. I don't think those readings in GPU-Z can be trusted as accurate, but that's just my opinion.

Maybe you could get Gamer's Nexus to test it out.

jollib
Can anyone confirm with a specific x570 motherboard they they are reaching 500w? I want to do a test to see if its motherboard related or card related. Ill pick up the same motherboard you said works and see if i can get passed 450

Double confirming my upgrade is an MSI meg x590 godlike.     I wonder though now if its temp related.  my card seems to never go over 73. My new case/setup is much cooler than my old one at idle, so i have to assume it's cooler all the way up, but it still seems to have a flat line at 73 degrees.  I have that temp slider in precision oc set all the way to the right, i wouldn't think it would limit by temp. but im not really sure?

Dabadger84
What sort of clocks are you guys getting out of your vRAM specifically?  Like what offset are you finding is stable?  Right now I'm testing at +195/+1100, and so far O_o it's no errors & score is still increasing.

I can set it to what you mentioned and it never crashes, but my fps and score goes down.
 
My FPS / port royal score is the highest at 105 core 205 mem. turning either up by even +5 and my fps and score goes down.

robotbeatrally

Dabadger84
What sort of clocks are you guys getting out of your vRAM specifically?  Like what offset are you finding is stable?  Right now I'm testing at +195/+1100, and so far O_o it's no errors & score is still increasing.

I can set it to what you mentioned and it never crashes, but my fps and score goes down.
 
My FPS / port royal score is the highest at 105 core 205 mem. turning either up by even +5 and my fps and score goes down.

I went through steps to ensure that was not the case.
 
Started at +120/+0 and worked my way up, +135/+500, +150/+500, +165/+500, +180/+500, then I saw someone reporting +1100vRAM, so I went back to stock on the GPU core & tested +750, then +1000, then +1100, none showed a score decrease, so I resumed my trek upwards, did +195/+1100, passed - went to +210/+1100 & it crashed midway through, backed back off to +195/+1100, closed all unnecessary applications in the background, ran it again, & got the 14728 score I now sit at for a "personal best".
Keep in mind these GPUs have clocks that scale at +15 increments, if you set anything in between, it could end up going up or down, and if it goes up when you would've been stable down, then you have more "work" to do to discover where you were actually stable, and where stability left the chat.

Dabadger84
 
 
I went through steps to ensure that was not the case.
 
Started at +120/+0 and worked my way up, +135/+500, +150/+500, +165/+500, +180/+500, then I saw someone reporting +1100vRAM, so I went back to stock on the GPU core & tested +750, then +1000, then +1100, none showed a score decrease, so I resumed my trek upwards, did +195/+1100, passed - went to +210/+1100 & it crashed midway through, backed back off to +195/+1100, closed all unnecessary applications in the background, ran it again, & got the 14728 score I now sit at for a "personal best".
Keep in mind these GPUs have clocks that scale at +15 increments, if you set anything in between, it could end up going up or down, and if it goes up when you would've been stable down, then you have more "work" to do to discover where you were actually stable, and where stability left the chat.

Those Clocks are on XOC BIOS right? For sure the Hybrid Cooler helps a lot because the problem starts when any of the sensors reports 60C+, under that with XOC BIOS on Air I can do great clocks.
 
I'm trying to buy one but EVGA doesn't accept international orders :/

Solved: 
My motherboard decided to downgrade the slot to x4 after installing my kit. Looks like after I gave it a good ole fashion blowout, it decided to run at x16 again. Scores are back to normal.
 
---------

 
I just installed my Hybrid kit, flashed the BIOS to make it a hybrid card, then flashed again to get it to the 3090 XOC Hybrid. 

I somehow lost 1000 points in port royal through all of that. Old high was around 13800, now I can't break 12800. What the holy f. Power never passes 410W full draw. 75, 110, 110, 110. Card will NEVER go over 100% power draw. :\
 

 
Edit: Going to flip the switch to Normal and see what happens. If anything I will revert the Hybrid bios back to FTW OC and see what happens.
 
Flipping the switch to Normal somehow is displaying the exact same BIOS version and Device ID. Wut? I never touched the normal BIOS ever. Same Port Royal score of 12800 ish.
 
Looks like I can't unflash it away from Hybrid and I am stuck like this for now. Tried all of the 3985 BIOS to get OC switch back to how it was and no go. Normal switch is busted I guess and just wants to act the exact same as OC :|
 
So with NVFlash and magic done, I was able to flash it back to 3985 OC then loaded XOC back on. I still am 1000 less than what I did in October (13919). I am capping at 12900ish using the same settings I had before. So I have no clue what is going on. I am never breaking 50C, average GPU Clock is hitting over 2040 more consistently. Just somehow my scores are 1000 lower. 

Difference from 2 months ago and now are PX1 from 1.1.2 to 1.1.3 and then the Nvidia drivers. Is there a known issue with drivers causing PR score loss?
 
 
 

Edit: wrong thread, thought I was responding to a different tab - mod delete double post please?