anvil

Well... for me it is the same when I see a 3DMark 2001 top 50 score with an almost unoverclocked GPU... Just have a look in the Hardware Masters category and you'll be astonished by the hardware those guys have benched. I cannot think this is their own hardware. Or if it is, they can afford a 32 CPU server easily.

To Wava: If you think that Hwbot users have to put only the results of their own hardware, waht are you doing with the people having 100 CPU-Z submission? Can you imagine the different platforms they used? Do you think it is possible to afford this hardware? And what about the people having more than 100 submission in all the 3D benchmarks? I really don't think you can compare hardware sharing and hardware from your familly or your job. And if you want to make the comparison how will you prove that this is not the hardware of the benchers? You will just apply the rule on the servers?

Hi, most of the time those results have been achieved using a highly overclocked CPU. Here we're dealing with CPU overclocking and not GPU overclocking, but the result is the same, better performances.

No actually some of the thorton were barton cores, and I had one. After removing the L2 cache it became a duron....

Hi, I also had this in the past, when I tried to add L2 cache on a barton. It turned out that I removed some cache: http://www.hwbot.org/result.do?resultId=567862

I do not think the money argument is making a point here. I really doubt that people can afford a cluster. I rather think that people are just seizing the opportunity that they have a cluster at work. I will also add, that having a skulltrail with two QX, a 1000W PSU, 200€ LN2 pots, liters of LN2 is far from beeing cheap. I do not know a lot of people that can afford this... We cannot split in two mono- and multi-socket systems because as previously said there is not a lot of multi-socket systems in Hwbot. If you create a new category for multi-sockets all the skulltrail will earn only few Hwboints and I don't think people is willing for this. And this is exactly the same for people having/using clusters they want to earn as many Hwboints as possible. The comparison with the GPU is not good because multi-GPU systems are cheap, so a lot of people will compete in the multiGPU category. Even a bi-athlon MP system is not so cheap, so no comparison IMO. I'm a little fed up of seeing always the same ones on the top of the ranking, bored when I see that people just keep their best score unsubmitted in order to beat the new WR in a two minutes time row. These are the rules, everyone can do whatever they want, but at least we see new faces and that really good!

This is exactly my point since several months now. HDD score in PCmark05 is now "over represented". But only now. In 2005 nobody would have think about SSDs, so at this time the HDD score was not too heavy, but it turns out that now it is. This is exactly the same issue with some 3DMarks and the CPU score. But no limitation exists for the CPU score in 3DMark. In my opinion the worst thing is that HDD score is limited in PCMark even though this benchmark is supposed to test the overall performance of the computer and the CPU score in 3DMark is not limited in this benchmark supposed to test the GPU performance... Do you see my point ?

I didn't know that Hwbot had a limit in hard drive benchs of PCmark05... I certainly have missed one news. What's the reason? And how hwbot users can perfectly fit this limit? Hard drives are not like GPU ou CPU (increment of 10 MHz really easy to make). What is the solution? Removing one hard drive from your RAID array? Or I'm also missing softwares allowing a limitation in hard drives bandwith ?

Ticket ID: 407 Priority: Low Hi, can you please add this videocard?\r\n\r\nhttp://img5.imageshack.us/i/fireglv7350.jpg/\r\n\r\nThank you.\r\n\r\nAnvil

You're absolutely right! Even the screenshot is almost the same (same wallpaper, and same windows layout).

Ticket ID: 356 Priority: Low Hi crew, \r\n\r\nI was looking at my scores and found an issue with this link:\r\n\r\nhttp://www.hwbot.org/quickSearch.do?hardwareId=GPU_1075\r\n\r\nThank you

Hi, Just give more time to the mods, they have a lot of job to do. At some point these scores will be moved to the right category.

I really not agree with thread crapping or pollution or whatever... If crapping was made, be sure that other words would have been employed. The main topic of this thread had just a shift, but at least it was alive and interesting! I explained to them some things because, even after their last BIG thread on the forum, it seems that JMHs don't understand Hwbot moderation (i.e. blocked scores, and the purpose of event notification beforehand).

Ok, thanks for the answer.

I know that VIP sections make the future of Hwbot, but only the CLOSE future. Perhaps I could have precise this point in my previous posts. The thing I meant is that this is not because Hwbot moderation is not thinking about it right now that they will never think about later. Do you see my point? I'm completely aware of the decisions that the moderation is making, and I participated to all the poll they made at the beginning of Hwbot. And their decision are fully fair for me. I'm not complaining about CPU sharing at all. It just makes sense to me that this will be a problem in the future. Perhaps I'm the first sharing this idea to the community, but when it's new, it's new, so it is not widely spread. I completely agree that this promotes teamwork, but it will (in the future again, not nowadays) also promotes teamextrapoints. Tweaks and customization are not physical elements so they can fully be shared. And hopefully they are shared. For RAMs and Motherboards I would say that they are not the limiting hardware, so sharing moderation cannot be applied to these components. @ Benji: You can only share items that you own. When you do not own it anymore, it is by definition not shared.

Ok, I didn't know about this... Too bad... What are the reasons? Cheating on Linux is really easy?

Hi Guys, in the next months (but I don't know exactly when) I might be able to run Wprime on a cluster based on several computers. This cluster is now running on Linux, but I think (but not 100% sure) I (or another guy from my work) will be able to install a windows emulator like Wine on it. What kind of validation can I propose in order to get the results accepted because I'm not sure the validation tool will work on this cluster. CPU-Z windows won't show the cluster processors, since I access this cluster through another computer, so CPU-Z will only show info about the computer I'm physically working on (it can be a Pentium s478). I'm a running jobs on the cluster thanks to a ssh command. Do not ask me more about this ssh thing because I'm a not the guy in charge of this cluster and I don't know how it works. The command /proc/cpuinfo will show some informations but not all of them I think. Any advice? Thanks

No offence but : The VIP section of Hwbot is not making the future as far as I know. With the evolution of CPUs, it will be possible to have World Record of some 3D benchmarks without the highest GPU. This is a fact. It has not happened yet but it will. Hwbot will have to deal with that. Suspicion is good for keeping this clear and fair, if suspicion was not present, you will have the 3D01 WR @ 300 000 3DMarks. Leaving the rules as they are have never been a good strategy. Just look at the recent issue concerning Wprime 2.0. If nothing had been made, people will still be able to submit scores with a non allowed release of Wprime. If including CPU in hardware sharing is ridiculous, I would say that more ridiculous is not to take into account that a CPU score can be more important than a GPU score in a global one. About "openly described as CPU limited benchmark by Hwbot (and others)": I just wanted to say that hwbot is not alone on earth. Just an theoretical example: If Hwbot is against one rule, and that all the overclocking community is defending this rule, I'm pretty sure that Hwbot will have to accept this rule to some extent. And now a real example: With R1ch we were the first ones criticizing the issue with Wprime 2.0 (http://www.hwbot.org/forum/showthread.php?t=2638). My first post about this problem has been made in november 2008 (http://www.hwbot.org/forum/showthread.php?t=2374). I'm not saying that I'm proud that R1ch and I finally succeed in making our ideas validated, because this would be totally wrong. But instead I'm saying that I am really proud that Hwbot moderation is able of criticizing its own rules, in order to make them better, because this rules are completely time-dependant. I don't know if my english is clear enough, but if one of you guys feel just one inch of insult in my sentences, you are making a mistake because I never meant this to happen. EDIT: @jmke : I've never said that such a rule was already existing.

I (and other people in this thread) never accused you of hardware sharing. In my sentences I was only talking about the worst case scenario. Concerning sharing CPU for 3D benchmarks of course it is allowed by Hwbot. Or I would say, "it is still allowed". I'm saying this because some 3D benchmarks are no longer 3D benchmark but CPU benchmarks. At the time of their release they were 3D benchmarks, but now after both evolutions of CPUs and GPUs, they are no longer 3D benchmark. Whatever will be the evolution of GPUs, the future Intel Processors will take the first place in 3DMark01, this is a reality. Hbwot crew will have to take into account this. And when those 3D benchmarks will be openly described as CPU limited benchmark by Hwbot (and others), the CPU sharing will no longer be allowed on those latter. So yes Jmax, it is allowed, but do it while it is still possible. PS: I am talking about future Intel Processors, but what if someone found a i7 capable of 5.8 Ghz? Will this guy be allowed to share his CPU to all his team members? I really doubt about it.

Unfortunately nothing can be done to avoid this kind of suspicion, or hardly. The best way I would say, is to submit scores at different times. Submitted at different times could mean benched at different times. But I don't think moderation will call you and Jmax as cheaters if you are submitting relatively close scores. Since these scores are not top 1 and 2 in global ranking I don't think there will be a problem. But if you are submitting top 1 and 2 scores, you will have to prove it very accurately. Because knowledge and tweaks can no longer be explanations for very scores at some points. To reach the summit you need really really overclockable hardware, and this cannot be shared as knowledge can be.

Exactly what I meant. If you have 10 CPUs and one can run 100 MHz higher than others, nothing prevents the different guys to use always the same CPU for CPU benchmarks or 3D benchmarks that are CPU dependant. But this kind of sharing is hard to uncover, except if each member of the team is submitting really high scores, all in a row. But people are not so dumb to do that.

Making such thread is just to prove beforehand that you had a lot of hardware in order to avoid some sharing issues. So if any sharing issue is made, you'll be able to prove that you had the hardware, and in this case your scores will not be blocked. But this kind of thread cannot prevent you from REAL sharing purposes. If the scores are too suspicious and if finally sharing is unveiled, scores will be obviously blocked and I also think that members will be (permanently) banned. If no suspicion at all will be made on your submissions, then perhaps you'll be the only one posting in this thread. Do not expect that the mods will say "Waouh! What a wonderful cooling or what a great score!", 'cause they here just to prevent some faked/cheated results from being granted with Hwboints, so that this international competition will stay fair as long as possible.

I just think they are now waiting for the benchs.

I don't get it...

I have two different things to say: First to Jmax: It appears that the rules have been posted twice, in two different news, at two completly different time periods. So telling that you or 99% of the community doesn't know this rule is not an excuse, and only shows that you and 99% of the community are not reading news. Second to the crew: If subset has only to be displayed by Hwbot members who have a high score thanks to a good tweak/opti, some of the JMH's have not to be deleted. Actually, I spend a lot of time looking at the results mentionned by JMH members, and they are really not so high compared to others.