Jump to content
HWBOT Community Forums

Recommended Posts

  • Replies 71
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Posted

Tbh, XTU is kinda a silly benchmark. It's just intel marketing, it only runs on intel cpus and it will always score higher on the newest intel cpu by a large margin. I think that maybe we should just get rid of the benchmark. If we just want a benchmark that uses newer instructions then x265 or time spy extreme should be plenty.

Posted
You do realise that xtu has been easy to cheat numerous times before but never had the old subs removed just updated to stop cheats

 

no had no idea, I just press the run benchmark button like anyone else should :o

  • Members
Posted

If there's no way of determining a real from a cheated score then I would suggest either removing it as a benchmark or at least removing points from old scores. And since hwbot doesn't want to anger Intel I would imagine the latter would be preferable.

  • Crew
Posted (edited)

well if everybody scores 1471, how on earth can you do 1480+? Is it automatically a cheated score? not sure...

 

Under cold XTU is way more consistant in output than on air or water, yet I rather see a lower point distribution and removal of most of them over the top subs...

Edited by Leeghoofd
  • Administrators
Posted

We might have to remove all top scores done with skl-x and also remove some other scores of different platforms that are "too good to be true". I am not sure disabling points for all xtu or disabling the benchmark makes sense. 99,9% of benchers did legit scores and only a few, maybe even some of those who already caused trouble at other benchmarks in the past, mainipulated scores. Why punish all when only a few did wrong?

Posted (edited)
We might have to remove all top scores done with skl-x and also remove some other scores of different platforms that are "too good to be true". I am not sure disabling points for all xtu or disabling the benchmark makes sense. 99,9% of benchers did legit scores and only a few, maybe even some of those who already caused trouble at other benchmarks in the past, mainipulated scores. Why punish all when only a few did wrong?

 

Some common sense has reached Earth !

 

We might wonder who needs to uncover this cheat now, to get some competitors doing a lot of XTU off the chart...? Probably a silly scenario of my mind watching too many movies...

 

taking down these scores will impact a lot of honnest folks. I am thinking about a Polish guy who would get blasted from the #2 or 3 if XTU goes down...

Plus many enthusiasts like myself, binning the best ram and board they can since at least that pays off in XTU..., then getting out the most their cpus can give, using the guidelines (i.e no cheats ; + I don't even run on W10 32 bits which everytbody think is the best...), and their "modest" material means !

Edited by bolc
Posted

Me too,seeing miker2ka benching another month xtu will be worth my money :))).

Leaving the joke on the side,if you disable or restart the points i think many guys will quit from this hobby.

Some maniacs out there stayed even 2-3 weeks for a score so imagine that you delete them... :))

 

 

Sent from my iPhone using Tapatalk Pro

Posted
Hi.

 

Instaled the latest version on Maximus V formula, and no system information is displayed on the right handside panel, plus whenenever I click on Run benchmark, error log is displayed.

 

Looks like this version sucks big time...

Can t recognize the cpu and number of cores... (i5 3570)

 

 

 

Got this message form another topic

http://forum.hwbot.org/showthread.php?t=162384

 

 

 

So I know 5.2 and 6.0 are ok with Ivy. 6.2, 6.3 and 6.4 are not. Will try 6.1 to see if it works

I can tell you in 6.1.2.13, the BM works, I have never tried submitting results but it does work (displays scores etc and all cpu info is correct). I have not tried later versions.

This is on Asus Z77Pro -i5-3570k.

Posted
Me too,seeing miker2ka benching another month xtu will be worth my money :))).

Leaving the joke on the side,if you disable or restart the points i think many guys will quit from this hobby.

Some maniacs out there stayed even 2-3 weeks for a score so imagine that you delete them... :))

 

Bad break.

 

Maybe HWB should think about actually enforcing punishment when the cheats are caught, instead of just slapping their wrists and sending them back in to do it again.

 

Non enforcement will cause more people to leave than 'selective' enforcement.

 

Unless that's the way HWB wants to roll, keeping their top 20 players and sponsors happy and losing most of the 'real' members.

  • Members
Posted

The issue isn't with crazy record breaking scores, it's with cheaters who actually think and just bump their score a little to get an extra place or two etc.

 

Especially with the no screenshot required etc

  • Administrators
Posted

The screenshot is something we should consider for sure, maybe in the way Albrecht proposed after rookie or Novice time is over and peope are deeper into the benchmark verification. We already made quite big progress at eliminating bugged runs, number has gone down drastically, but needs to lower this even more, to zero at best. This is the main problem we face that people still can talk about bugged runs when they cheat, we face this at other benchmarks as well btw

Posted
The issue isn't with crazy record breaking scores, it's with cheaters who actually think and just bump their score a little to get an extra place or two etc.

 

Especially with the no screenshot required etc

 

Completely agree with this. This is especially an issue with locked chips, where a few XTU points could be worth an extra 30 hardware points. How can you distinguish between someone who spent the time to strip and tweak the OS/Mem/etc. then reran for two weeks compared to someone who dialed in a couple of reasonable looking mem timings, then cheated their way to a 1st place score inline with everyone else?

 

Especially considering how valuable locked XTU is(50pts per sub), that could unfairly give someone a huge boost in the rankings.

 

On the other hand, if any legitimate subs get removed because they were "suspiciously efficient", that's just as unfair to that individual.

 

If there's no way to determine without any doubt whether or not a sub is legitimate, then the only fair way to resolve the situation is to reset all points IMO.

Posted (edited)
I vote for adding a proper screenshot too, maybe just excuse the Rookie class from that. Makes moderating a bit easier too

 

I voted for it a year ( or 2 ago ) and nobody even seemed to have read it lol.

 

but finally it is mandatory.( i hope ? )

Edited by subaruwrc
Posted

As some guys said and also I said in my previous comment, can't make exceptions... Killed HWBOT Prime cause few guys cheated? Do the same with XTU, same treat for all cases if you want to act professionally

Posted

Just to make this clear as it seems people are riling each other up in this thread: there is no indication of any significant amount of cheating in XTU (yes, we could spot irregularities in the cheated submissions)

 

We will NOT remove points for the >99.999% who play by the rules because of actions of the <0.001%.

 

I also want to say thanks to @RULE for bringing the problem to our attention prior to it going public. It gave us a much needed head start to address the problem before the sensationalists came along.

 

Sent from my A0001 using Tapatalk

Posted
It gave us a much needed head start to address the problem before the sensationalists came along.

 

I personally find this a very derogatory statement considering most people here are just rallying to have the problem fixed.

 

And in the usual HWB fashion, it will be done half assed.

Congratulations on your consistency.

Posted (edited)

We will NOT remove points for the >99.999% who play by the rules because of actions of the <0.001%.

 

Well, this is what you made with HWBOT Prime... Why threat them different?

 

Edit: This case is even worse, potential points gained are way higher than with HWBOT Prime

Edited by Alan_Alberino
  • Administrators
Posted (edited)
Well, this is what you made with HWBOT Prime... Why threat them different?

 

Because we can fix the loophole on XTU which Intel did and we cannot fix the hole at Hwbotprime up to date - simple as it is

Edited by websmile
Posted

 

Especially considering how valuable locked XTU is(50pts per sub), that could unfairly give someone a huge boost in the rankings.

 

On the other hand, if any legitimate subs get removed because they were "suspiciously efficient", that's just as unfair to that individual.

 

If there's no way to determine without any doubt whether or not a sub is legitimate, then the only fair way to resolve the situation is to reset all points IMO.

 

Can't be said better than this...

Posted
Well, this is what you made with HWBOT Prime... Why threat them different?

There is no inconsistency.

 

X265 points were restored when we found a way to 1) spot cheated submissions and 2) the developer was able to identify a fix quickly. This is also the case for XTU. The difference between both cases is that for XTU we were able to start moving BEFORE it went public. I was in touch with Intel the day we received the news and in a call a couple days later was reassured a new version was in the works that addresses the issue.

 

For HWBOT Prime this is sadly not the case.

 

(Note: after we found a way to spot the X265 cheats, the staff went through a whole lot of submissions only to find 1 person had been doing this)

 

Sent from my A0001 using Tapatalk

Guest
This topic is now closed to further replies.

×
×
  • Create New...