speed.fastest Posted February 17, 2016 Posted February 17, 2016 So this is not a benchmark, but score based on value you set that detected by XTU but must finish the test. What make me think this is not a benchmark because variable score is too tight, even if i test XTU with Youtube opened, score is not affected Quote
crustytheclown Posted February 17, 2016 Posted February 17, 2016 Some initial feedback on this topic. This issue was highlighted to the XTU development team on May 7, 2015. Our report was based on findings of user @henkenator68NL who provided us with the test data via @HiVizMan. Based on the findings, the XTU development team added a fix in v6.0.2.8 which prevents a user from changing the priority of the benchmark process or any other XTU related processes. This version was released in October 2015. Based on this thread we'll check if this issue is still present on the latest XTU release and if so, check with the XTU development team for more information. Nope you can change every process' priority. They just moved all the processes in a temp file that gets erased at the end of each benchmark. It's very easy to just copy these temp processes while running xtu and change anything you need to them and then just throw them back in the temp file to be ready for the next run.Simple as that. Quote
Guest 0.0 Posted February 17, 2016 Posted February 17, 2016 New dual core record http://hwbot.org/xtu/share/359012?clientVersion=6.0.2.8 Yep, still needs improving. (The bench, not the score ) Quote
Massman Posted February 17, 2016 Posted February 17, 2016 New dual core record http://hwbot.org/xtu/share/359012?clientVersion=6.0.2.8 Yep, still needs improving. (The bench, not the score ) Thanks, can you send the data-file to pieter@hwbot.org? A "how-to" would be nice too Quote
Splave Posted February 17, 2016 Author Posted February 17, 2016 Thanks, can you send the data-file to pieter@hwbot.org? A "how-to" would be nice too how to is in the first post Quote
Guest 0.0 Posted February 18, 2016 Posted February 18, 2016 Thanks, can you send the data-file to pieter@hwbot.org? A "how-to" would be nice too Just did a quick install and run, didn't save any datafile. There's http://hwbot.org/xtu/settings/359013 if that's any use. I don't think one can ever stop some few people cheating, all we can do is hopefully make it hard enough that they will not try. By posting methods it makes it easier for them and just opens a can of worms IMHO. Better the development team/person overhauls the bench as the title says rather than just dealing with the posted vulnerability. Quote
Administrators websmile Posted February 18, 2016 Administrators Posted February 18, 2016 I saw too much results out of line to think that this is new to some people, opinions can vary on if it was OK or not to publish this, but personally I refuse to blame the messenger, though it is always good if you contact staff first. "Tweaks" like these spread fast in the dark, at least now we have the awareness of people on this. Very tough situation, let´s hope there can be a fix Quote
Doug2507 Posted February 18, 2016 Posted February 18, 2016 Sure, junk it and stop chasing the money trail. Quote
Massman Posted February 18, 2016 Posted February 18, 2016 Just did a quick install and run, didn't save any datafile. There's http://hwbot.org/xtu/settings/359013 if that's any use. I don't think one can ever stop some few people cheating, all we can do is hopefully make it hard enough that they will not try. By posting methods it makes it easier for them and just opens a can of worms IMHO. Better the development team/person overhauls the bench as the title says rather than just dealing with the posted vulnerability. Thanks, forwarded the info to devs. Quote
Administrators websmile Posted February 18, 2016 Administrators Posted February 18, 2016 Sure, junk it and stop chasing the money trail. It is an easy to do bench even for newcomers, it is supposed to give hints on system performance and show people that benchmarks and measuring system performance can be fun. Not all is about the money, and if you know me, you also know my opinion on big cash dominating the ranks P.S. anyway, we need a fix or we have a constant problem^^ Quote
dumo Posted February 18, 2016 Posted February 18, 2016 If easy to cheat then the rest will be sooo easy Quote
Guest george.kokovinis Posted February 18, 2016 Posted February 18, 2016 My 0.02 cents in this thread. Hade never heard about this Lasso thing. As a matter of fact, I do not know anything about apps intervening in benchmarks. I read about Wazza and LOD tweaks and it is like Asimov stories. Any way, I decided to download this application and give it a shot on Z170. Followed Splave's instructions to the last comma. Well, it appears that this thing has no effect on Z170. More feedback from others with Z170 is a must, because maybe I do not know what I am doing ( I would not be surprised ). By the way, Latest XTU version is installed on Win10 x 64 Pro. Quote
Spiedie Posted February 18, 2016 Posted February 18, 2016 Just tried on Z97 W7 with HPET enabled and disabled: that seems to be the issue. With HPET enabled and Process Lasso enabled, p95-bench.exe runs at normal priority. After disabling HPET and rebooting, Lasso does its thing and it runs at Realtime priority (causing the issues). Well, it appears that this thing has no effect on Z170 If skylake is unaffected then is looks to be the same old RTC bug (or Intels attempt at fixing the issue with RTC in XTU, fixes breaking things never happened before right?). Quote
Doug2507 Posted February 18, 2016 Posted February 18, 2016 You can run lasso real-time with hpet. There's 2 files that need priority assigned for full effect. Quote
zeropluszero Posted February 18, 2016 Posted February 18, 2016 It is an easy to do bench even for newcomers, it is supposed to give hints on system performance and show people that benchmarks and measuring system performance can be fun. Yeh but so is every other benchmark. Quote
Doug2507 Posted February 18, 2016 Posted February 18, 2016 Give em 3d mark instead. At least it's got pretty pictures... Quote
TX_OC Posted February 18, 2016 Posted February 18, 2016 As a quick to program, temporary fix until a proper fix is released, xtu devs should make xtu compare core frequency with max achieved frequency Quote
FUGGER Posted February 18, 2016 Posted February 18, 2016 Some insight: Intel has contracted a 3rd party to code XTU since it was moved from "Iron City" The programmers name is Ben and he is close to the Intel engineers I work with. George, Allen, and I have met Ben. Ben took this picture: https://www.facebook.com/photo.php?fbid=1612698442343092&set=a.1612698452343091.1073741828.100008088543899&type=3&theater If we can figure out a fix to make Ben's job easier that would help, suggestions are welcome. Massman and I know who to contact to escalate this. Quote
Massman Posted February 19, 2016 Posted February 19, 2016 You can run lasso real-time with hpet. There's 2 files that need priority assigned for full effect. Can you PMe the exact instructions so we can replicate and patch? Thanks :celebration: Quote
Doug2507 Posted February 19, 2016 Posted February 19, 2016 Can you PMe the exact instructions so we can replicate and patch? Thanks :celebration: Sure pj. Away to do fresh os install today so I'll check out what the other .executive was. Maybe Allen could tell you if he's already setup? P95.exe and can't remember the other. Quote
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.