https://overclock3d.net/news/software/pro_overclocker_uncovers_3dmark_performance_cheat/1
 
Futuremark has been made aware of this issue and is likely to release a fix for this problem in the near future.

Weird.... I just said that timespy results could be flukes back in November because of garbage results where an entire system can be slower but score better, even though all results point to a much better faster system, and low and behold a glaring issue is found and reported.

How the hell could someone load different drivers in the validation time is beyond me, unless they are doing offline results. All I know for sure is that timespy is a garbage benchmark when it comes to legitimate results.

I reported an issue to them a while back that was a glaring mistake on their end. Faster cpu’s We’re showing to perform significantly better, but the results always said the slower cpu outscored the faster on the comparing scales. Thankful futuremark was quick to repair that, but how the hell it went unnoticed is beyond me. For this that didn’t know, the physics test goes by how long it takes to complete a physics task, so it runs in milliseconds. A 5960x takes about 80ms where a 7980xe takes about 30ms.. the compare results were in favor showing the 5960x faster, because the numbers were bigger so it had to be better... once they realized it, they corrected it right away.

One reason I could care less about Time Spy.
Sad though that someone who Benchmarks has to find a cheat to boost his scores.
Some are so Hard Driven to only see the Graphics Score and not the Overall Score as well but that is just how it is.
 

I don't understand , did not everyone know of this ?? 
 
LOD gets you a better score   what's next he'll find the power slider j/k  
 
btw , you don't have to wait for the benchmark to start running , when using nvidia inspector do not apply it to the global profile open the drop down select TimeSpy , hit apply and enjoy flatland every time it opens if you like it that way.  Also it only works on Time Spy to any degree , the rest of the fire strike family does not respond very well if at all.  

Don't use the global profile select Time Spy in the drop down then apply whatever setting you like , max performance , single display , ect
 
 

Minor changes good for an extra 300 points. Nothing wrong with that, lol.

Ok Sam we are gonna get this straight right now.

I am in my early 50's and I don't cheat ... I don't need to.  Cheaters don't post a How To Guide on how to beat their scores
 
I was showing how Rubass's discovery is not news to everyone and even more that he's doing it 1 way and I can reproduce it another way , I also posted this at HWBot to him so he can see there's more ways than his to get a valid result. If nothing this might help the guys at UL figure out how to fix it faster. 

And for the record if you want to compete at HWBot http://hwbot.org/user/sam_nelson/   I Highly Suggest you start READING THE RULES for competition. I did not make them but I do FOLLOW them if you don't like it ? Take it up with HWBot.  http://hwbot.org/user/hillguy/  

LOD in Time Spy is considered a CHEAT ( it has been since NOV 2016 ) but in every other benchmark it's an allowed tweak / optimization.
 
 

 

 
What say you now Sam ??
 
 

Hillguy
Here's a comparison with and without the "tweak" in Time Spy ...  in order of apperance in compare link.

no lod  https://www.3dmark.com/spy/3203906 .... 9713
lod on global profile http://www.3dmark.com/spy/3203952 .... 9908 INVALID RUN
lod latest version of inspector http://www.3dmark.com/spy/3204076 .... 9906
lod older version of inspector http://www.3dmark.com/spy/3204033 .... 10018

Compare link https://www.3dmark.com/co...py/3204076/spy/3204033

I just ran this all back to back with no special timing to start the test required so I don't know what he's on about there.    

Older version of inspector is ... ? Thanks

Cool story... I think?

Honestly, if you don’t care about benchmarks, why comment on a thread about them saying you don’t care? Seems like a waste of time.

I admit, I am guilty of doing the same thing, which is why I have stopped replying to threads that I actually don’t care about. Saves time.