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

Lol, yeah, I figured the results here on [H]ard is but a fraction of the overall number of results. I did a reboot and managed a much better score however.

Graphics Score: 37,777

https://www.3dmark.com/3dm/66335313?

On Air

Still getting the Hardware monitoring disabled message. Will need to look into that further.
 
Lol, yeah, I figured the results here on [H]ard is but a fraction of the overall number of results. I did a reboot and managed a much better score however.

Graphics Score: 37,777

https://www.3dmark.com/3dm/66335313?

On Air

Still getting the Hardware monitoring disabled message. Will need to look into that further.
I almost jumped out of my chair when you posted this. I just assumed it was a single card. Then looked at the details and realized it's SLI.
 
cool least its not just me for a while i was going crazy trying to figure out why i was getting lower scores on my 5950x vs my 3900x i had before.
If it was a just a cpu swap on the same hardware, did your memory settings change in the bios? You can get a healthy boost on Ryzen chips if when your memory is dialed in to the sweet spot for whatever your motherboard can handle. What type of an overclock did you have on the 3900x vs the 5950x?

Rebar should theoretically not show any substantial difference between the two different cpus unless you were running on different motherboards.
 
reminds me i need to redo some of my benchs i got 64gigs of ddr 3600 running 4000 now stable had to switch to pbo tho from a ccx overclock so working on tunning that also.
 
Double check your effective clocks. Its a really good score but you might be overshooting the OC leaving some time running at the higher clocks on the table. I personally don't like to see my effective more than one speed bin off from what I'm aiming for.

I'm not sure if I could get requested and effective that close together at these high voltages and clocks. I didn't check that time, but based on what I've seen on previous runs it was probably running ~2170 effective.

What I've been doing is starting with a slightly steep VF curve and then flattening it until I find the max effective clock that can complete the run. That seems to give me finer control since I can manipulate effective clocks in steps less than 15mhz.
 
21k on my 3080ti a few weeks ago... I'll be dropping in a 9900ks in a week, might bring the score up a bit.

3DMark Basic Edition_2021-10-19_13-08-45.jpg
 
Anyone have a good comparison of the launch drivers and bios vs the current drivers and/or bios comparing 3090s with identical settings/setups, to identify the driver side progress nvidia may have incrementally achieved so far this gen? I think we're far enough away from launch that this is likely going to be a good representation of what Nvidia was able to better optimize one year out.

Do you believe Nvidia still defaults some very specific settings to acheive a better score in futuremark benches while toning down some of the 3d overhead? I continually notice some of the settings seem purposely scaled back, designed to default for weaker visuals, lower gpu resource time, and better performance.
 
Last edited:
Anyone have a good comparison of the launch drivers and bios vs the current drivers and/or bios comparing 3090s with identical settings/setups, to identify the driver side progress nvidia may have incrementally achieved so far this gen? I think we're far enough away from launch that this is likely going to be a good representation of what Nvidia was able to better optimize one year out.

Do you believe Nvidia still defaults some very specific settings to acheive a better score in futuremark benches while toning down some of the 3d overhead? I continually notice some of the settings seem purposely scaled back, designed to default for weaker visuals, lower gpu resource time, and better performance.
I dunno, this thing has so much ass behind it I'm not sure I have even noticed any changes. I play all games fully maxed at 4K... lol. I DO have benchmarks from around when I got the card vs. things now (I always save the runs to compare). However, I think my Timespy score has remained in about the same ballpark, give or take a % which a simple reboot could cause as well. I'd probably bench even better right now with the cold weather! :LOL:
 
The best PR driver for me is still the older one that I ran above. Not a huge difference, but I usually get 50pts or so more with it. Not sure if it makes a difference in TS, etc...
 
I recently upgraded to Windows 11 and a 12900k asus hero MB, with corsair vengence ddr5. on

my 3090 had a crash during an msi OC Scanner, now my 6 month only 3090 went from ~19,000 to ~16,000. Is this a windows 11 issue, something hardware wise with the new or is my GPU failing. How should I diagnose?
 
I recently upgraded to Windows 11 and a 12900k asus hero MB, with corsair vengence ddr5. on

my 3090 had a crash during an msi OC Scanner, now my 6 month only 3090 went from ~19,000 to ~16,000. Is this a windows 11 issue, something hardware wise with the new or is my GPU failing. How should I diagnose?
I would pop a win10 install on a spare ssd and just test that. You might try to dual boot as well, though I don't know how much fun that is with 11. You changed a lot at once, but I think a quick os install is the least painful thing to try.
 
I would pop a win10 install on a spare ssd and just test that. You might try to dual boot as well, though I don't know how much fun that is with 11. You changed a lot at once, but I think a quick os install is the least painful thing to try.
Well I dunced out and forgot I could just slap her in my old system so thats next.
 
https://www.3dmark.com/3dm/70189098?

I think my score might be limited due to my 34" lcd having a 100Hz max refresh. My next upgrade will be a faster lcd, been waiting for the panel quality to get better in the fast refresh types. 240hz seems a good target.
 
Still no updates :( , is there a way to give everyone access to edit a post? Maybe just quote the first post and change?
 
Back
Top