r1ch

Ticket ID: 834 Priority: Medium Hey guys,\r\n\r\nCan\'t seem to edit one result of mine:\r\nhttp://hwbot.org/community/submission/958493_r1ch_pcmark_2005_core_i5_661_29022_marks?tab=more\r\n\r\nWanted to add the GPU and clocks I was using. \r\n\r\nCheers,\r\nRich

Congrats Jody, great score.

thanks dude

Congrats Nick! Don't keep us waiting on 6s!

Wow Nick, great time! You laughed at me when I suggested sub-6 seconds 2 weeks ago!

Before it decended into a flame-fest... My thoughts too. This whole thread is a little desperate, in my opinion.

03 is about as close as you get to a 3d benchmark now. With rev 3, people now focus on one card so there is less of a cpu bottleneck which is a good thing that many don't realise.

Just affects Clarkdale or Lynnfield as well?

Nice score mate.

how many more backups nick?

If points are the only reason to attempt a world record then I think this is very sad. Compare this to footballers who get paid £100k a week to play football because they are the best in the world for their clubs but then refuse to play for their country at the world cup "because they don't get paid enough". On the development server, kingpin dropped to 4th I think. He benched 1x5870 quickly and that's the only reason he's #1. It will take time for everyone to do this - not everyone can react so quickly. You are riight at the end, the best overclockers - people like you, hipro, kingpin and andre are still there at the top of the leaderboard!

I don't think this change is fair for all types of users - this is going back towards Rev2. Why can't Rev 3 be given a chance to settle down? At the moment the single GPU categories are very hit and miss for the quality of results. Once there are more results in the single GPU categories, it will take much more skill to rank highly in the single GPU ranking than to bench 5 pots. Just because a user can manage 4 cards, and 5 pots, doesn't mean they're not being rewarded in hwpoints - they can prove their skill and earn their deserved points with one of the cards - this is the Rev 3 competition. The single card category is where overclockers skill can be fairly evaluated without money or sponsorship getting in the way. This is what hwboints are for and the reason Rev 3 was introduced, was it not? This is the whole point of Rev 3 which has barely been given a week to settle in. WR's are a significant achievement, yes I completely agree, but I don't think they should be given more points (in terms of the pure ranking for differentiating who's a better overclocker). Maybe this is the perfect opportunity for Achievements to be promoted more. "Ruled the rankings: 3DMark01", "I was the king: 3DVantage" etc? Maybe there should be a "World Record" medal that appears next to someone's name when they've been skilled enough to take a world record? I think this is the road we should be going down. If this must go ahead, I would like to see it delayed for a week, a month, something to let the Rev 3 settle in a get a better idea for how the points distribution works out and the skill required to get those points. I would also like to see the points greatly reduced, and the number of positions reduced to the top 3, or maybe top 5. Those points are only achievable to the single card categories though, I think? I think hwbot is in a difficult place - you are trying to give people who don't have massive sponsorship and endless money the chance to compete, but trying to keep those people who do have that happy as well. I don't envy you the decision!

Gratz Nick, loving the mods to the 5870.

I take nothing away from your previous score, we've all spent time, money and effort getting scores that have now lost a lot of points, but that was always going to happen with the change. You have got to give the current system time to settle down. If you benched a GTX 280 or 285 in the same way you did before, you would score lots of points. We should be looking into the future, seeing what is the best for future scores, not what makes past score 'fair'.

While I am all for rewarding world records, I think the current Rev 3 already does this. This "extra" points total would take things too far. For example, take the Aquamark #4 postion, 1x HD 5970, stock cooler, not overclocked. http://hwbot.org/result.do?resultId=925857 Scores 29 points already, with this change it would score 110 points I think that's too much.

How much of a rule is this? I've seen a couple of submissions where they're not the same, and then a couple where they're different but probably the same person. I don't know if it falls under this 'grey' area that it's only if there's suspicion?

I have tested this tonight, a number of tests score a lot higher. Can this be blocked please? I am sure it is an honest mistake by Jody. Another one: EDIT: reported the normal way http://hwbot.org/result.do?resultId=900564

First, I'm not sure if I'm right with this so I don't mean any offense to 3oh6 and if it's not a problem that's fine. http://hwbot.org/result.do?resultId=860974 PCMark05 default resolution is 1024x768, his res is 800x600. Does a lower res give a higher score in PCM05 like 3DM?

The 220 limit was previous applied by futuremark when submitting to get an orb link for Top #20. If they still have the limit then it's implied in the rules by saying the top #20 need a valid orb link. If fm have removed the limit, it should be mentioned in the hwbot rules.

Hi jmke, what are the problems with splitting into single socket and multi socket categories like the 3D benchmarks? I think this would be a good idea. If we know what the issues are, then hopefully we can find a way of solving them.

In my experience, OS plays a huge part in wPrime, and there are further tweaks that you can do to get better scores on top of that. I'm not saying that every time is perfect, but if there are suspect times, how about working out the PP for them (like Superpi) and seeing if there are any real anomalies? I would be happy to do some testing with my tweaks to give a ballpark "PP" figure.

Gothrek, you're fighting a losing battle and I doubt anyone will agree with you. Memory sub-timings can have a huge impact on a superpi 1m run, let alone a 32m run. To dispute that is just crazy.

Thanks for replying jmke, thanks for asking rb and I like your thinking on the faster time I know I asked before, so I'm sorry if I'm repeating myself but you didn't answer the question in my last post...Can you say whether it's ok for me to modify my time 0.01s faster so that I get the correct points and rank? Yes, it would have the wrong time, but it doesn't affect global points, and it would be "getting around" the 2 dp problem. As knopflerbruce says... Can you answer these suggestions please?

Bump?

jmke, I'm not going to do anything that hwbot doesn't want me to do - as you've explained, hwbot only works to 2dp. What I'm asking for is special dispensation from hwbot to make a change that gives me the position and points that I should have already. It's only a 3rd place Q8200, and 0.5 points we're talking about, so it's no big deal - it's not like it's a world record!