Jump to content
HWBOT Community Forums

Deliberately disabling GPU Cores for 3D Benchmarks


Recommended Posts

Guest hammertone
6 hours ago, FireKillerGR said:

Jokes aside there needs to be a distinctive line to whats punishable and what isnt and to what extent.

Recall @Raufgetting banned cause he altered a screenshot of a legit score as he added a memory tab that was missing.

I call it " play fair kiddies "

Your parents if you behave bad. What do they do? Take away your toys.

If you are a grown up now. What then? One person in the group behaves bad. Then everyone gets punished. Expect it has always been that way.

Lose your points? Someone is to blame. Lose your loved 3D01se? Then that has me shitty. Then I am hating on someone real bad ?

Link to comment
Share on other sites

Guest hammertone

Would not like my dictatorship.

Loop hole wizardry would get 1 month lock out of Hwbot and it's forum. Make you think twice about your exploits. 

Lucky eh ? You have kind moderation team here

Link to comment
Share on other sites

Well ,  it took me just 30 seconds to find the outlier score talked about here  ? . It would be a sad day when 01 and 03 benchs are gone , those are 2 very tweakable benchs for single and multicard at a driver level and other legit tweaks to make them scale like @saltycroissant said , please do not let one isolated incident of one person take the fun away. Maybe a wrapper like Aquamark maight be possible or something to detect when running multi gpu.

 

 If all else fails ,  i suggest Alby @Leeghoofd buy a beer  ? for everyone who reports a legit cheater ? .

 

 Happily Retired From Puerto Rico with Love - Angelo 

  • Like 6
  • Thanks 1
Link to comment
Share on other sites

  • Crew
6 hours ago, Mr.Scott said:

Only a month? You are more generous than I.

In the good old days the silver hammer was applied, I think remember that...

But my friends this case is not the only one as mentioned above.

There are already at least two more reported cases and not only Splave, Bullshooter I think has been reported for the same reason in 3D01

I think we should ask the big overclockers to stop doing this , and delete your incorrect 3d01

Stop Now, please

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

Guest hammertone
8 hours ago, Mr.Scott said:

Only a month? You are more generous than I.

eh el-cap-i-tan ?

A name and shame leader board of past perpetrators 

Link to comment
Share on other sites

On a side note, when I was benching multi-card setups, I knew that setting catalyst AI to disabled was actually switching off CFX - never exploited that thing yet results would easily pass video check as GPUZ and driver would report CFX is ON from start to end of the run (and I guess even for futuremark systeminfo it would be flagged as run with crossfire enabled).

So it was always a problem for bencher to decide if it is unfair or not - me personally was thinking of it as unfair.

Link to comment
Share on other sites

I understand why 01 multi-GPU might be killed at this point but eventually this would result in the same punishment for those who spent the money & time to have it figured out properly & fair with the one(s) that cheated.

Additionally, what is funny is that the one(s) who cheated on 01, couldn't get proper scores on 03 cause there the scaling is much higher/more evident.
This results in your "3x/4x" runs (single gpu) to be far lower compared to the actual/proper runs.

That alone should be a dead giveaway that others weren't cheating; it was just yourself failing to figure things out and turning to cheating as the solution.

Edited by FireKillerGR
  • Like 7
Link to comment
Share on other sites

If anything this thread demonstrates that the community can be relied upon to spot when something is wrong with scores that get big points, and raise a stink.  Obviously only scores getting big points get community attention, but that would be the case however moderatable the benchmark is.

The rules page used to have a "disallowed tweaks/cheats" section.  Bring that back and put the onus on the benchers to report any of their own scores where they might have got a bit overzealous, and to stop taking the wee in the future.

Then, if people are still cheating, leave it to community members with a personal interest to figure out what they think is wrong and explain it.  Leave it to the person accused to show how they got the score and prove it is legitimate.  Leave it to the community to accept the proof or point to a problem with it.  Mods shouldn't have to run around collecting evidence, it shouldn't be much more difficult than checking a verification link.

 

It's easy to fall into the trap of holding others to the same standards to which we hold ourselves, maybe I'm biased because I was willing to send a board and CPU across europe just to remove doubt when @unityofsaints on my then-team had a score in question.  But if you're claiming to be fastest in the world, backing that up with a video doesn't seem like an unreasonable ask.  Not like people are saying 01 runs need an o-scope or anything ?

  • Like 7
Link to comment
Share on other sites

I vote for removing globals for any multi-GPU categories that do not scale with GPU-power. It's just artificial points.

Would we keep a CPU benchmark that only scales with GPU? We already have enough points for the highest clocking CPUs...

  • Like 2
  • Confused 1
Link to comment
Share on other sites

Well if we loose 01se..... imo the most fun benchmark because of all the tweaks, can we at least give Vantage or 05 globals for multi gpu. So like Stavros said we don't screw ppl who took the time to optimize and get those damn hawaii card to behave. 

Took me a  whole month to figure out what was wrong with my driver setting ?

Edited by saltycroissant
  • Like 2
Link to comment
Share on other sites

3 hours ago, Rauf said:

I vote for removing globals for any multi-GPU categories that do not scale with GPU-power. It's just artificial points.

Would we keep a CPU benchmark that only scales with GPU? We already have enough points for the highest clocking CPUs...

Im with Tobias here. Same thing for 05, 06, Vantage, 11. All of them score lower in multi GPU than single GPU. 

From HWBot perspective, any benchmark that requires extreme tweaking is bad for us. It puts high workload on the moderation team and is at the same time demotivating for people who want to step into overclocking. 

1 hour ago, saltycroissant said:

Well if we loose 01se..... imo the most fun benchmark because of all the tweaks, can we at least give Vantage or 05 globals for multi gpu. So like Stavros said we don't screw ppl who took the time to optimize and get those damn hawaii card to behave. 

Took me a  whole month to figure out what was wrong with my driver setting ?

That's why this is no option from my side.

  • Like 2
Link to comment
Share on other sites

  • Crew
6 minutes ago, der8auer said:

Im with Tobias here. Same thing for 05, 06, Vantage, 11. All of them score lower in multi GPU than single GPU. 

From HWBot perspective, any benchmark that requires extreme tweaking is bad for us. It puts high workload on the moderation team and is at the same time demotivating for people who want to step into overclocking. 

That's why this is no option from my side.

Roman shouldn't globals be added to other benchmarks to compensate for the loss of those globals ?

 

Link to comment
Share on other sites

Thats why i proposed other benchmark because in fact vantage does scale by adding more gpu's just like 3d11. ADL is fast enough to keep up with the overhead of quad gpu or triple gpu. For a clear exemple look at my x4 3D11 sub my cpu isnt at 7G yet im graphic score is much higher then everyone else. 

  • Like 5
Link to comment
Share on other sites

  • Crew

We are looking into if we can  drop globals for the multi GPU part of eg 3Dmark 2001SE. So not all globals are lost, we want to at least keep some globals for the Legacy benches. Not everybody benches latest gen CPUs nor GPUs.

@Sweet 's idea is spot on , the clockers concerned should report their own scores and the moderation Team will remove them. No further drama required. It is not that one got away with it that you should do exactly the same

 

  • Like 3
Link to comment
Share on other sites

18 minutes ago, der8auer said:

I don't think this helps. Different benchmarks will be lead by different people. You will not get an equal "compensation" in any scenario. 

Well thats why i proposed vantage because r9 295x2/12900k will still lead there. The only inconvenience im seeing is having to rebench but...imo thats better then loosing 200-300 globals with no possibility to get any back with that same hardware.

  • Like 4
Link to comment
Share on other sites

Guest hammertone
4 hours ago, Leeghoofd said:

@Sweet 's idea is spot on , the clockers concerned should report their own scores and the moderation Team will remove them. No further drama required. It is not that one got away with it that you should do exactly the same

 

Correct. Best option.

Then no one else gets affected.

Link to comment
Share on other sites

Guest hammertone

HWbot has a good idea recently. Adjust points upwards so average overclockers can make more points 

and feel good about their efforts. So you keep the remaining die hard overclockers interested to submit scores.

Some of these members have bought 3 or more of same graphics cards

Done the old benchmark correct. Happy with their effort ?

__________________________________________________________________

Remove benchmark is penalty. Punish 99% of members who do right thing. Enjoy 3D01se. Who are happy doing score.

That is loss of money + loss of time + loss of points. Why? Punish all? Because of the few.

______________________________________________________________________________________________

Leave our lover 3D01se out of this. Benchmark has done nothing wrong. It is the user who is at fault.

Get to the root of the problem. A few top overclockers here " only " pushing the limits. That is moderation team place to focus.

Link to comment
Share on other sites

1 hour ago, hammertone said:

Get to the root of the problem. A few top overclockers here " only " pushing the limits. That is moderation team place to focus.

that's essentially how 99% of the rules were created. There is always a grey zone and if it's not forbidden clearly, people will abuse it.

  • Sad 1
Link to comment
Share on other sites

I still struggle to see how removing or adding gpu to sub in a categorie that isnt the amount use, to be a grey zone.

 

This might sound salty..... but bullshooter cheated, there no other way around it. We shouldnt have all that useless debate because he's been here for a long time....

He cheated- receive appropriate sanction.

The End

 

Id expect the same to happen to me if someday i decide that screw that im not making the effort to understand a benchmark im just going to cheat.

Edited by saltycroissant
Link to comment
Share on other sites

To be honest, the efforts of the "new owners" of HWBOT are not new to let Multi GPU die...

image.thumb.png.dd480a1225b8ec7589698fc16c5d7bc8.png

 

 

I have spent a lot of time trying to get this bitchy setup to work with these AMD cards - not to mention that it is very difficult to get working cards at an affordable price. Testing different drivers, driver settings, monitors, refresh rate settings took me days. I felt the same way as Saltycroissant...

However, it has to be said that the AMD driver offers a good opportunity to disable Crossfire without even knowing.

The driver creates an application profile for each benchmark. In the "3dmark01" profile, "crossfire" is deactivated in the standard profile, although it is activated in the global profile and GPU-Z also shows this. So you should either delete the corresponding profile or set the "crossfire" option manually...

The evidence is then the corresponding values in the "Nature" run.

 

In 3D Mark 03 you have no possibility to get better values without crossfire if you are limited with the CPU.
Anyone who claims otherwise is simply wrong - sorry.

Single GPU - 1x R9 290X @ 1290 / 1700  -  259530 (CPU @ 5.82 / 8P)

https://hwbot.org/submission/5005275_oviz_hardware_lab_3dmark03_radeon_r9_290x_259530_marks

4x GPU - 2x R9 295X2 @ 1175 / 1550  -  367550 (CPU @ 5.82 / 8P)

https://hwbot.org/submission/5001509_oviz_hardware_lab_3dmark03_2x_radeon_r9_295x2_367550_marks

 

If you cheat here - you cheat yourself. I do all this just because it's a hobby for me and I am very ambitious to achieve my results under the given conditions and according to the set rules!

I am really pissed off...

 

Edited by OVIZ Hardware Lab
  • Like 6
Link to comment
Share on other sites

I dont bench multigpu a lot, but dont remove them, i enjoying seeing how the score created and some of them like 3DMark11 for Physics score actually scale a lot from memory more than cpu! You can try that! Im 3DMark11 big fan here, but i like 01, 03, 05, 06 and Vantage too, so dont kill them.

  • Like 4
  • Thanks 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...