RTX 3090 [H] Owner's Official 3DMark Time Spy Leaderboard

I finally installed my 3090 fe and got around to doing some 3D mark runs.

I came from a 3080 and saw my voltage go up to 1.081 during benching. When I ran the 3090 it was jittery and below 950 mv. I figured that the 3090 just ran at a max of 950mv and 3080 at 1.081v.

I just finished a bit of modern warfare and found out that the 3090 does go up to 1.081v with higher clocks to boot.

How come I can’t surpass 950mv in benching while I can go up to 1.081v in gaming?
 
I finally installed my 3090 fe and got around to doing some 3D mark runs.

I came from a 3080 and saw my voltage go up to 1.081 during benching. When I ran the 3090 it was jittery and below 950 mv. I figured that the 3090 just ran at a max of 950mv and 3080 at 1.081v.

I just finished a bit of modern warfare and found out that the 3090 does go up to 1.081v with higher clocks to boot.

How come I can’t surpass 950mv in benching while I can go up to 1.081v in gaming?
power limit
 
power limit

so there is no way to get above 950mv in benching? I maxed the power slider all the way up to 114%. I can only manage 1980mhz in port royal whereas war zone and cyberpunk I can get 2150.
I didn’t think power limit was different depending on the application.
 
so there is no way to get above 950mv in benching? I maxed the power slider all the way up to 114%. I can only manage 1980mhz in port royal whereas war zone and cyberpunk I can get 2150.
I didn’t think power limit was different depending on the application.
you could try overclocking. You don't really want higher voltage, you want higher clocks. Each GPU has its own limits.
 
so there is no way to get above 950mv in benching? I maxed the power slider all the way up to 114%. I can only manage 1980mhz in port royal whereas war zone and cyberpunk I can get 2150.
I didn’t think power limit was different depending on the application.
You can find a bios that will work with your card to raise it. You can explore your boost curve to see where you get the best OC vs under-volt.
 
Thank you both.

I don't feel like changing the bios at this time. I'm bummed that I'm stuck at 14200 in port royal, but atleast I can do 1.0v at 2080core in gaming
 
Thank you both.

I don't feel like changing the bios at this time. I'm bummed that I'm stuck at 14200 in port royal, but atleast I can do 1.0v at 2080core in gaming
If you want a higher score, overlock the gpu and ram. The game is not as intensive as the benchmark so clocks can go higher (way oversimplified). Put the fan to 100, open the door or window if you’re really crazy, and increment the card by 15mhz all the way up until it crashes PR.

it sounds like you’re just not happy with performance vs. other 3090s but I’d say your card Is pretty good if you’re at 2150 during gaming.
 
If you want a higher score, overlock the gpu and ram. The game is not as intensive as the benchmark so clocks can go higher (way oversimplified). Put the fan to 100, open the door or window if you’re really crazy, and increment the card by 15mhz all the way up until it crashes PR.

it sounds like you’re just not happy with performance vs. other 3090s but I’d say your card Is pretty good if you’re at 2150 during gaming.

So far, my RTX 3090 Founders Edition hasn't clocked any higher than 2,115MHz during benchmarking. It won't do much more than that as that's been about the limit of what it can handle while gaming. I'm on watercooling as well.
 
So far, my RTX 3090 Founders Edition hasn't clocked any higher than 2,115MHz during benchmarking. It won't do much more than that as that's been about the limit of what it can handle while gaming. I'm on watercooling as well.
Lucky ... My EVGA 3090 FTW3U won't go above 1950 with the 500w bios. Totally power limited.
 
20540

Asus TUF OC - stock bios

This card seems to want to put a hard wall up at 65*c for power limits. It's so touchy!
 
New high for my Zotac. Still on air, max fans. This is being stupid and flashing the KP XOC bios on it for a while. ~500w. Obviously need better cooling although core never got over 70C. Not putting that back on until I do the custom loop.

https://www.3dmark.com/spy/17900454

I thought I had a screenshot of GPUZ but apparently not.

CPU score is borked because I have SMT off for MSFS right now.
 
its winter in the uk.. maybe 6°C outside
fan was on 100% ofc

Its not even that good tbh if i took the side of the case off im sure it would drop a bit more
as airflow around the gpu isnt that good in my meshify s2
 
its winter in the uk.. maybe 6°C outside
fan was on 100% ofc

Its not even that good tbh if i took the side of the case off im sure it would drop a bit more
as airflow around the gpu isnt that good in my meshify s2
Did you put your PC outside for the run? I haven’t seen a 3090 FE on air under 50°C on Time Spy and 50°C was only achieved when they were undervolted.
 
Did you put your PC outside for the run? I haven’t seen a 3090 FE on air under 50°C on Time Spy and 50°C was only achieved when they were undervolted.

Check out my score. 41°C avg using A/C. The FE is simply a beast if you can feed it cool ambient air.
 
Did you put your PC outside for the run? I haven’t seen a 3090 FE on air under 50°C on Time Spy and 50°C was only achieved when they were undervolted.
I was able to get down to 49C with my 3090FE on my last competitive run that was submitted here. Aside from running the fans at 100%, no special tricks were done. Case was closed, no open windows or doors, house thermostat set to heat and around 22C. I was on the leaderboard for a while but it looks to have recently been clipped off.
https://www.3dmark.com/spy/16230285

Since that last run, I honestly haven't made any new attempts to see if I could squeeze more out of it. The settings I was using were mainly just for 3DMark and not something I am using for my day-to-day use.
 
I was able to get down to 49C with my 3090FE on my last competitive run that was submitted here. Aside from running the fans at 100%, no special tricks were done. Case was closed, no open windows or doors, house thermostat set to heat and around 22C. I was on the leaderboard for a while but it looks to have recently been clipped off.
https://www.3dmark.com/spy/16230285

Since that last run, I honestly haven't made any new attempts to see if I could squeeze more out of it. The settings I was using were mainly just for 3DMark and not something I am using for my day-to-day use.
sorry bro, cut you from the list last night, there were a lot of new submissions. Eventually the list will all be 22k runs :D

EDIT: bank error in your favor. You're back on the leaderboard.
 
Last edited:
Check out my score. 41°C avg using A/C. The FE is simply a beast if you can feed it cool ambient air.

I was able to get down to 49C with my 3090FE on my last competitive run that was submitted here. Aside from running the fans at 100%, no special tricks were done. Case was closed, no open windows or doors, house thermostat set to heat and around 22C. I was on the leaderboard for a while but it looks to have recently been clipped off.
https://www.3dmark.com/spy/16230285

Since that last run, I honestly haven't made any new attempts to see if I could squeeze more out of it. The settings I was using were mainly just for 3DMark and not something I am using for my day-to-day use.
Both of these scores I didn’t question because one was on A/C and the other is 49°C which isn’t an outlier. But 44°C on room temperate ambient air makes me raise an eyebrow.
 
I'm half tempted to try the 500W bios for my FTW3U to ensure I can keep my name on this list... lol. :LOL:
 
Both of these scores I didn’t question because one was on A/C and the other is 49°C which isn’t an outlier. But 44°C on room temperate ambient air makes me raise an eyebrow.
you and me both, but I'm not here to police testing methods, if someone lies about their test setup to get higher on the board, that's their own problem. Not calling anyone out, just stating a fact.
 
sorry bro, cut you from the list last night, there were a lot of new submissions. Eventually the list will all be 22k runs :D

EDIT: bank error in your favor. You're back on the leaderboard.
Woohoo!

Hey, I'm just surprised I've managed to stay on the board this long, especially with just a stock 3090 FE. I knew that once some of these higher-end AIBs and custom BIOS cards with higher power levels, my days on the board were going to be numbered. It's still a bloody fast card, even at stock settings.
 
Woohoo!

Hey, I'm just surprised I've managed to stay on the board this long, especially with just a stock 3090 FE. I knew that once some of these higher-end AIBs and custom BIOS cards with higher power levels, my days on the board were going to be numbered. It's still a bloody fast card, even at stock settings.
TBH, I think is all comes down to keeping it cool for higher boosts and silicon lottery. I think there is enough data out there to show that there is minimal difference at best between all of the cards out there.
 
TBH, I think is all comes down to keeping it cool for higher boosts and silicon lottery. I think there is enough data out there to show that there is minimal difference at best between all of the cards out there.
I somewhat agree. Cards with two 8 pin connectors will roughly perform the same. Their bin will dictate their performance more than the smaller power limit differences. But cards with three 8 pin connectors (with 450W+ power limits) should show a bigger increase over the two 8 pin cards as long as they have decently good silicon.
 
TBH, I think is all comes down to keeping it cool for higher boosts and silicon lottery. I think there is enough data out there to show that there is minimal difference at best between all of the cards out there.
If you keep the card cooler it uses less power and doesnt ride the power limiter. There is quite a few differences in cards having a large window of power limit differences and there is differences in how well you fair in the silicon lottery. FE cards power limit at 400w while most others are within the 450-520w range.
 
Both of these scores I didn’t question because one was on A/C and the other is 49°C which isn’t an outlier. But 44°C on room temperate ambient air makes me raise an eyebrow.

I tell you what bro when i get up in 12 hours
i will reinstall 3dmark and take my side off
and see what she goes down to with the back door open
 
I tell you what bro when i get up in 12 hours
i will reinstall 3dmark and take my side off
and see what she goes down to with the back door open
You don't need to rerun the test for me, I honestly don't care lol. Was just interesting to see a 44*C temperature while on air without any mention of additional assistance (frigid air from the outdoors) to bring down temps in the original post.
 
Evga 3090 FTW3 Ultra with 107% Power Slider and +150gpu +1000mem. Air Cooling.

Got some better case flow going and seemed to help my score a little bit. Gotta make sure I stay on the leaderboard with this card darnit... lol.

21445 Graphics Score

https://www.3dmark.com/3dm/57893940?
 
Updated score with side off:p

RTX 3090 FE Air
i9-9900KS

Graphics Score 22 218

https://www.3dmark.com/spy/18096908

1.jpg2.jpg
 
Check out my score. 41°C avg using A/C. The FE is simply a beast if you can feed it cool ambient air.
Honestly I think most of the cards respond the same - due to Nvidias power/thermal management. My Asus card will do essentially anything it wants up to 65*c, then it just manages everything to keep it at 65*c or lower, despite my protests. If I opened my window for the night (it has been in the negative *f lately) I would imagine that it would be full gas, based on everything I've seen so far.
 
Ok, switched to water. Nothing fancy though. Just 360mm radiator with 1500rpm Noctua fans. At this power level the core temp was 59C at the end of the run anyway. Water temp goes up quite a bit. Can't get a very high OC out of the Zotac which is holding it back quite a bit. It'll go higher without crashing in TimeSpy, but I know for a fact it will crash in Port Royal or Cyberpunk if I go much more. So this is a bit more legit.

XOC bios
~550w
21922 graphics

https://www.3dmark.com/spy/18127692

Quitting there before it blows up. LOL
 
Back
Top