the_Scarlet_one
Trying my highest run ever right now. Results if it passes in a moment.

Nevermind, lol, “clock watchdog timeout”. Not sure what that means, but I think I found the limit lol!

Nice scores. 

Fastest run yet: 4.817ghz 5960, 2088 Titan Xp.  Freezing cold in my room.  Hands hurt, A LOT!  (2100 mhz failed first run.  closes program to let the GPU get to 10c before I try again)
 

Snipes7
New run http://www.3dmark.com/spy/3328037
6192
 
Cheers

That CPU score is utterly ridiculous!! Lol! Dang!

Are you going to grab a 7980xe and try it as well? Curious how 4 more threads will help on chilled water.

A quick run on the new version with the same settings as my old run. Note: new version has Meltdown-Spectre firmware updates and they do effect the Physics score.
New version: http://www.3dmark.com/spy/3329152
Old version: http://www.3dmark.com/spy/2763626
 

https://www.3dmark.com/3dm/25235191 (2152/1600) 

Snipes7, if you haven’t read the forums over there, the product manager commented back:

Pasi Virtanen, 3DMark product manager
Hello Clemens,

Apologies for none of us getting back to you on your email - while it spread through the team, the responsibility of getting back to you with the end results of our investigations was lost in the process as your new submit also triggered some other internal investigations into the new patched version.

This round of invalidations was a special case due to the then undetectable way of cheating spreading to media with reports of abuse while the window of opportunity was open.

For your Time Spy score in particular, it ticked the boxes consistent with the publicized cheating method compared to the mass of about 10000 recent other single-GPU Time Spy Extreme ones (including looking at monitoring data of clocks during the workloads which didn't fully explain this and also lower scoring similar results from other EVGA cards with the same IDs), but now that the dust has settled, we've determined this to be a false positive for this case and we have reinstated your previous score to the Hall of Fame.

Apologies again for the communication drop. I will again stress that this was a highly unusual measure taken only due to the widely reported and publicized method of cheating.

For anyone else with similar issues, feel free to email me with details at pasi.virtanen@futuremark.com and we'll take a closer look at a case-by-case basis.

- Pasi Virtanen, 3DMark product manager

Snipes7
Oh nice! No I didn't read that. Gotta love justice I guess I take back / apologize for what I said in the heat of the moment.

No problem. Hopefully there won't be anymore false positives.  

Snipes7
Yes you are right, it's valid again! Well it was in the HoF... until I got a new score http://www.3dmark.com/spy/3330611
 
@CptSpig mine are also on the new meltdown patch... in 3Dmark it's not that big of a hit... check out XTU, it's insane, like at least 20% if not more. I guess I have to go back again.

Yes that is true for most Futuremark runs. For TSE it uses AVX instruction so the physics scores are effected a little bit more than normal. Edit: I forgot to change my AVX offsets to 0 I think that's why my physics scores were low. As with the XTU and Cinebench R15 I do notice a drop in my scores. Can't reach this now but I am working on getting back with different settings. When I get a chance I will work on getting my TSE scores up a little.
 

 
 

It shows your card as a collectors edition!!!! Can you send me your bios, I want to see if it makes a difference lol.

Well how do you cool your titan? It might just be Temperatur :)

Nicklas0912

the_Scarlet_one
It shows your card as a collectors edition!!!! Can you send me your bios, I want to see if it makes a difference lol.

Cause it are a Collectors Edition
 
send me a pm :)

Something I dont get, the first day I got my card into the system, I get 10500 GPU score in timespy "normal" stock clocks, and after I resintalled windows it drop by 100, I did reinstall 3 times to see, but it did nerver come back to 10500..

It is probably the 3% margin of error in all FutureMark benches. 100 points could also be you have something running in the background or you loaded a different Nvidia driver. If you keep running the program without letting the machine cool your score will drop. Unless you have a chiller to keep water temperature constant.

Do you have in the NVIDIA Control Panel the setting Set to Max Performance?

Nicklas0912

bcavnaugh
Do you have in the NVIDIA Control Panel the setting Set to Max Performance?

Did not help at all :) even got the 3% lower score with the max perforamce on.

Then maybe you have another odd issue, that setting it where you want it for Benchmarking.
3% From what Overall Score to Overall Score? Check Each Score to see what score went down.

Thought I would add mine... https://www.3dmark.com/spy/2620634 for TS Extreme. First run. Mild OC.
Regular TS: 1 Card. Pretty decent OC. https://www.3dmark.com/spy/3187587 (My second best single GPU run)
Regular TS: 1 Card. Highest stable OC. https://www.3dmark.com/spy/3187683 (My highest GPU clock, but due to CPU getting too hot and throttling, the score is lower ultimately)
Regular TS: 2 Cards. Basic OC (2012 x2) and a high / moderate CPU OC. https://www.3dmark.com/spy/3353894
Still haven't played with my second GPU enough to locate it's potential. I plan to push it hard and water cool both cards, and not just my CPU as these cards put off a lot of heat collectively.
 
-Jacob B.