I.M.O.G.

Thanks Sebastian!

Thanks guy. I did do it before dolk... although it didn't turn back on the next morning. I haven't tested the chip again, but it is quite possibly dead. So dolk never got to take his own run at it. I had tried getting 275x25.5 for 7013mhz, but couldn't get a validation so settled for this one. So we compromised and drank crown with the guys.

BTW, if you want to search for scores on a specific piece of hardware, just start typing in the search box, then wait for a list of options to drop down... It was a bit slow for me to populate, and you don't get the advanced search options available on rev3, but it works.

Leave it blank. There were optional fields on Rev3 you didn't have to fill out also.

+1 Searching for past benchmark scores is the first thing I do before a session, to set my goals for what I'm looking to beat. Does it exist and we just aren't finding it?

Good stuff Bob! Benching party is in Philly next month, hope you can make it this year.

Thanks Miah! I bought some RAM to improve it, however unfortunately it wouldn't work in either of my AM3 boards, and now I've tested it on P67 and it doesn't work in that either - the guy I bought it from said it didn't work for him on 1156 either, but worked fine on 1366. The quality of my ram was the only thing standing between me and the efficiency I needed to change this 10th place score into a 4th place score. In 1024m, the ram wasn't holding me back: http://hwbot.org/community/submission/2136980_i.m.o.g._wprime_1024m_phenom_ii_x4_965_be_3min_3sec_499ms

Thanks dude! The people in my livestream were pretty helpful.

Just wprime1024 and 32 runs at these speeds, my superpi's are a bit faster but could be improved also. Still haven't figured out the problems with pifast, pcmark05 3d pixel shader, and 3dmark06. 3dmark03 runs fine, as well as wprime1024. Tried at stock and room temps, as well as -190 at a low 5Ghz. Missing something, but can't figure it out.

Nevermind, I'll just post something higher. I'll take this as a sign! :-D

How do you request hwbot staff correct points? This is worth points, as I don't have any higher 1090T submissions.

It has come to my attention that there has been some disagreement amongst hwbot staff over these accounts. As the operator of Overclockers.com, and having met Dejo once in person, I am interested in resolving any uncertainties. My understanding is that similar submissions, and the timeframe on submissions, has aroused uncertainty regarding if both accounts are legit. Is that what the problem is perceived to be from hwbot staff? If so, please inform me of any materials you require. Proof of purchase for separate GPU for 3D and separate CPU for 2D can be provided. Dejo's real name is Jonnie, and his daughter is October. He volunteered to host the Overclockers.com LN2 Benchmarking party last year. He is an asset to the benchmarking community, and an avid HWBot overclocker. October became interested after hanging out at the party with us. Jonnie then took the appropriate steps to do everything the right way... 1. He asked Massman about using his old 920 for her 2. Massman took a while to reply, Jonnie bought a new 920 for her in the meantime 3. Jonnie is teaching her the structure of benchmarking and overclocking, and their submissions look similar 4. He has 2 GPUs, but has only used one for her and one for him. 2x GPU submissions have been made only under one account. Cards are labeled so they know who's is whose. All screenshots are stored in separate folders on a USB stick Basically, I only met Jonnie once, but thats all it takes to know he's a straight shooter. He's trustworthy, honest, and he doesn't like bullshit - he'd be the last person to try and skirt the rules. His daughter's account was clearly named so it didn't look like they are trying to get away with anything, transparency. So maybe it doesn't count for much, but as the operator of Overclockers.com, I have no reservations about personally vouching for the legitimacy of all their submissions and him following the letter of the hwbot law - I'd put my own name and reputation on the line. October has used her own GPU for 3D, and her own CPU for 2D. That is what the rules state. Please let me know what materials are required in order to demonstrate these are 2 separate accounts and legitimate people. Here is a picture of Dejo and Mrs. Dejo: Here is a picture of Dejo's Daughter, October, sitting in the middle of the table with a blue bracelet. This was taken at our family dinner on the last night of the benchmarking party: Just tell us what we need to do, and we can clear up any misunderstanding. I can be reached at 440.499.4221 or matt at overclockers.com. We'll provide any proof or materials requested. Dejo is a good dude, and its a cool thing to spend time with a kid, so I just want to help straighten this out.

Voted no. There's always some level of uncertainty, but the current system is working well. If you look at enough scores, its not hard to tell what is reasonable and what is not. Confirming the motherboard info does little to ensure the scores are legit, but is a bit of extra work. Sounds like more worrying about cheating instead of worrying about our own benching. I vote for more focus on benchmarking and the love of the game, and less worrying about what other people are doing. The obvious cheaters and the questionable submissions will be scrutinized, and if some slip by then so be it. If you are working hard at your own benches, there aren't enough cheaters or invalid submissions to make a big impact on you. If you encounter something questionable, report it. Let the judges and mods worry about it. Otherwise, focus on your own benching and thats the key to success.

When submitting scores, its fairly easy to leave the score field blank on accident. Maybe I'm the only idiot around that made this mistake, but a lil javascript would keep people from hitting submit without filling out that field. Please delete this at your convenience: http://hwbot.org/community/submission/2099769_i.m.o.g._wprime_32m_phenom_ii_x4_965_be_0sec_0ms

Please delete. Sorry.

This was submitted by a moderator at Overclockers.com. He is a good person, and an honest overclocker. Please be more specific if there is a problem with the submission that needs corrected. However, it is not a "fake".

Has HWBot considered reporting statistics on the number of active members on a team? For example, on the following page it says OC Forums has 262 members who have contributed to the team total points: http://hwbot.org/community/team/oc_forums?tab=members However, that tends to lend the impression we are a huge team, when in fact we are a relatively small team with a long history. Every person currently active on a consistent basis knows the rest of the guys personally, and we consistently get a few members joining up every month to kick in a few points but not all of them stick around for the long haul. It'd be cool to see some insight into how many teammates our team has active in the past 30 or 60 days maybe, and compare that to other "big" teams.

Just a heads up in case it wasn't on your radar, but the custom flag for OC Forums is broken since the update also. Its trying to display the image from this url, but its no good: http://hwbot.org/signature.img?iid=253617&thumb=false Other custom flags like this one are working fine: http://hwbot.org/signature.img?iid=318544&thumb=false Is there anything we need to do, or do we just need to wait for you to finish fixing things up? No biggy either way, just a heads up in case you weren't aware.

Actually mutant, it doesn't work for sigs - while technically it would, many browsers don't operate well when there's a thread displaying many posts and each users sig is running the hwbot javascript. The browser gets laggy and runs like crap with that much javascript, we've seen it happen in posts on OCF when we were first testing the feature as R_B was developing it. But voidn is on the right track for something with a lighter footprint, we do something similar with folding stats: http://www.overclockers.com/forums/showpost.php?p=6558662&postcount=6 If we were grabbing the XML and pumping the data into a DB table, it wouldn't be much of a leap to then implement this ourselves for benching stats: http://folding.extremeoverclocking.com/?nav=IMAGES The largest problem with doing that is using PHP to generate the actual text/image - getting the layout right is similar to making a really complicated vbulletin sig, it takes a lot of trial and error to get right.

I wanted to let you guys know I like the direction you are moving in, the consideration that is apparent with how you've planned, and the way you are aiming to move forward. As for the feedback you receive, I support your confidence in doing what you believe is best for the community - regardless of who it pleases or who it upsets. At the end of the day, you are the people who have been running the most successful community based overclocking competition platform out there - you are the experts. Here are a few thoughts I had posted previously, regarding some of the community concerns with the issues you are aiming to address: I think its clear hwbot is making the right decisions for the right reasons, and as you move forward I think the outlined plan should be a big success - regardless of who is happy and who is upset. Overall, I think you are moving in a direction that is an improvement and overall the community will be positive about it. Thanks for the work you are doing on things!

The way I see it hwbot.org is the best competition platform there is for the consumer PC hardware industry, despite its shortcomings like this thread highlights. If you distill all the complaints and compliments, the bottom line is that most everyone competing is doing this for fun - even the people getting the ES chips are doing it because they like playing with hardware. The people complaining are more worried about their boints as they compare to someone else's than they should be, and they are losing sight of the fact this is fun even if the playing field isn't 100% level with scientific accuracy. I voted for option 1 because in the end the boints are just numbers systematically assigned to give us something to compare eachother with and have more fun. If you want to compete with people at the top, you have to play the game a certain way and position yourselves the way those individuals have, simple as that... If you don't want to play the game that way, there are many other ways you can compete on hwbot and still have a lot of fun. I definitely see the value in looking at ways to improve and standardize better, but I think massman and richbastard have struck a pretty good balance with the way things currently are. There will always be griefers regarding the details, but the way submissions have grown and subzero has exploded... That's a testament to the fact these guys are doing a lot of things right here and a lot of people are enjoying it. And to the griefers - if you guys motivate people to look for better solutions, in the long run you are helping move things forward too and thats great. Its good to better understand the perceptions people have and find whats important - if hwbot doesn't do it, someone else might. In the end tho, I think the right ingredients are here and they'll keep improving in the future as they have in the past. So Props from Overclockers.com - I may not be an active bencher, but I have a lot of fun keeping track team OC Forums and watching them compete against the other teams and members around them. Thanks for the great platform to do it upon hwbot!

As has been discussed, the contest did not end when it was requested to. This was a limitation of the competition platform at hwbot, and outside the control of Overclocker's Benching Team. Regardless, I apologize to those who were affected. We will be more aware of these sorts of complications, and will take good faith steps to avoid them in the future. We appreciate hwbot's assistance and efforts while hosting this competition! Thank you hwbot, massman, and rb for your help! Winners and entries are being evaluated based upon the existing rankings listed on the contest page from 1st to 10th place in each stage. Prize pack winners will be announced officially through Overclockers.com, very soon. We'll update this thread at that time also. Also, please note that the points displayed for submissions on each stage are incorrect. For how points are assigned, and how winners are determined, reference the rules announced officially here: http://www.overclockers.com/overclockerscom-benching-team-superpi-competition/ Those rules are also listed correctly on the hwbot competition page here: http://hwbot.org/competition/ocf_spi_comp?tab=rules Points summary: 1st place in a stage gets 10 points, 2nd place gets 9, and so on until 10th place gets 1 point. Points in 1M and 32M for that platform are added together for each contestant's total score on that platform - 1156/1366, 775, or AMD.

Thanks for breaking the ice Sam!

That's pretty funny! For all participants, be sure you use the official contest background released here when submitting (to help confirm your results were obtained during the competition): http://www.overclockers.com/superpi-competition-under-way/ Direct link: http://www.overclockers.com/wp-content/uploads/2010/04/OC-First-Annual-Super-Pi-Competition-Official-Wallpaper.jpg