NVIDIA gimping benchmarks: GTX 780 tested {322.21 vs 353.62 vs 376.33 vs. 398.36 vs. 411.70}

Or maybe you should've clearly expressed in the OP that you're using your own benchmarking method instead of official build-in tool that was used everywhere else including guru3D
I can say I'm getting 200 FPS in firestrike 4k and in a week casually mention that I mean not average but max in one specific scene. This "Fraps 15 second benchmark was used" could mean that you used it to test games that don't have a dedicated benchmark, for all we know. Stop overreacting, I didn't say all your results are bullshit, I did however agree that GTX 780 can't produce 72 FPS average in DX:MD build-in benchmark. That's it. If anything, I'm glad that @norton could instantly catch differences between widely known results and yours, that's why I liked (and still like) his post. Peace.

 

It goes the other way too, users with modern GPUs believe that their cards are hampered by the drivers supporting the older series.
Back when the DX10 cards were moved to legacy, and Fermi this year, users seemed to believe their modern cards would get some magical speedup by removing the code for those cards.

 

Sorry for the misconception then! The official benchmark, does indeed, stress out GTX 780 to quite low performance numbers. Getting 28/35/45 FPS on ultra settings and 42/50/63 FPS on high settings. So then, what do those numbers mean in the Guru3D benchmark? 30 FPS on ultra settings and 42 FPS on high settings - are those minimal, medium or maximum FPS?

 

It's an average, since it's stated that internal game benchmark was used.

 

When someone reports benchmark results, usually it's from the in-game benchmarking tool. That's first and foremost, it should go without saying. Countless benchmarks are performed this way. And here you are, launching the game - flicking the mouse a few times and reporting a 'benchmark result' of 72 fps. And then expecting an apology when people point it out.

What was it again?

Do it right, or don't do it at all. I wouldn't have a problem with it, but your attitude makes it worse.

 

416.16 did have worse perf:

 

I forgot about the benchmark simple as that. Typically i try to do the benchmarks with built-in ones. I said i am sorry that i forgot about the official benchmark! Same goes to you @vbetts, what did you say to me? crap posts? Is that an official word form a moderator? I was trying to to defend my position that i did not fake the results, and i proved that! I showed 72 FPS in a typical gameplay recording screen, having forgot about official benchmark, and you called that shitpost? Can you repeat that? Your manners are poor for a mod that's been for a millennia. I liked this forum for a long time, despite not actively joining conversations, but that attitude of yours calling my post a crap really drops me off to visit the forums further. Don't make Guru3d reputation worse!

And to the point: if Guru3D scored 30 FPS on ultra preset and 42 FPS on high preset with GTX 780 Ti, that means the benchmark was done somehow messed up. I scored 35 FPS average on ultra preset and 50 FPS average on high preset, and with a weaker GTX 780 no less!!! Not only that, but GTX 780 Ti scored a lower number than R9 380X, what? In what kind of GPU world does that nonsense happen? And i guess you were fine with those results and did not question them?

So what was that again? You can't be serious and show that benchmark result as a reference point in 2018. This test should be redone i think, those number do not make sense.

 

no it didn't.

 

Looks like it did to me. Unless that bench is fake, of course. But that youtube channel seems to be rather legit.

 

GF 411.70 is "my system" best driver ,most stable for window 1809 .

 

Spectre nerfed that driver.

its been irefutably proven by lokken

 

Do you have a link handy? Also, does that mean 416.34 is fine?

 

https://forums.guru3d.com/threads/nvidia-slows-down-performance.423434/#post-5595670

 

Yeah disrespecting a staff member will get you on everyone's good graces. Plus the mod was doing his job. You were being disrespectful towards other members. And also you were posting misinformation as well. Some one posted actual evidence calling out the video in your original posting by posting the benchmarking results done by Guru3d. At the end of they day Nvidia is not gimping the cards. It's the games becoming too new for older cards to handle. The GTX 7xx series cards are 5 years old at this point. Plus I reported you for disrespecting a staff member.

 

That's realistic.

Just stay calm don't get upset ...

When you did your DeusEx Benchmark run and received the results you did you need to consider your results could have differed based on the following:

- Different CPU / speed of
- RAM / speed of
- Chipset
- Driver revision tested on and how refined it was compared to the baseline your testing against ( ie GURU3D)

When you completed your DeusEx run unless you had "like" variables in a "like" run in the case of trying to "replicate" something you can't discount theirs for instance. It doesn't mean they did something wrong. If this is the line of work or hobby you want to be known for these things shouldn't need to be pointed out to you they should be self-evident to you at this point.

Normally I would've been more "kosher" in my choice of words but you strike me as someone with self-inflicted arrested development, sir/ma'am/kid'.

 

Those of us who have been using Nvidia cards for more than a decade know that these urban legends are false, so I consider that this topic should be closed because it only contributes to:
1.- To feed unfounded and invalid myths.
2.- Gives visits to channels of doubtful quality.

 

No, it doesn't.
You don't have to believe everything that comes out on youtube.
Be more thoughtful and reason more before you believe some youtuber.

 

Yes of course, i am glad we all have settled and don't need to argue. Sometimes the built-in benchmarks are not as good as people might think, because they don't recreate the same results each time and provide big variables, which should never happen. This is not the case with Deus Ex, however, it's a great benchmark, even though it kinda shows the worst case scenario, because in a typical gameplay, in ordinary levels, the performance of Kepler seems to be much better then what the benchie shows.

 

Now that's just stating the obvious and missing the point of what I was trying to relay to you. I'll quote myself.

Let me dumb it down to simple English as much as possible. I was calling you out for saying GURU3D's results were bunk without having shown your own like comparative in a like environment with like variables to offset their results, as is the case with any form of reproductive test ie be it Scientific experiments etc. If you have the ability to attempt to disprove something - do it.

Talk is cheap. I've said all there is to say to you. If you don't understand complex thought, articulate discourse or proper sentence structure that isn't my problem nor am I being paid to be your Tudor. Good day Sir/Ma'am. *Christ.... the people.... *

 

It's "tutor", not "Tudor", which is a dynasty of English monarchy. If you are going to gripe at someone about their abhorrent grasp of the Queen's language, you should apply the same zeal to your own typings.