knopflerbruce

You should use the category of the best card, and then count the number of cards you're using (1 5970 + 2 5850 = 3 cards), so pr def. you should set it to 3x 5970 - however such a combination doesn't exist, so the correct category is 2x 5970. The engine doesn't work very well with this type of multi GPU setups, it's not an ideal solution - but it's better that a handful of submissions get less boints than a shitload get too little because of one of these setups

Bumpetybump. I suggest we use the names suggested by Tapakah (with MHz after the FSB numbers), but maybe add the rev. number too, in case people separate these chips that way. I can move scores, but I want to be sure I'm doing the right thing - can you confirm this, Karl?

Nah, not that much - was more like 370 I think, plus a couple of globals (!)

Thanks No outstanding scores, but a few boints here and there is always nice:)

There's a burger king by track 14 at the train station. What more do you need, seriously?? Double steakhouse FTW!

It's gonna take a while if it has to be done manually, and I want some more MOD opinions on this matter before I do that (if I have to).

no need to keep the other account, as it's empty Dunno why it was created in the first place.

I actually agree with K404. There is very little difference (at least on the extreme end of the rankings) between the two cards, and since it's not possible to tell what card it is, perhaps merging is the best solution here:)

Fixed,

Then most likely you'll only be able to use one CPU when you bench:p

True, but what if the CPU or GPU was bought on a local auction site, and it turned out to be the same chip as someone else used? This is not uncommon I think.

I'm thinking it's a bug, but on the other hand... the scores do actually beat the ones below them, and AFAIk he's not using a special sample - so in some way, it's fair enough that those results "steal" boints from the worse scores. However, I didn't code this - Massman must clarify if this is intended or not:D And if it can be fixed.

I don't see how it is relevant that some guy benches with Hipro, but the FS part is fine - it's not forbidden to sell a piece of HW within the team, as long as it's not being passed around like it's everyone's property.

The user needs to recalculate all his best submissions, then it will reappear (if it's not still blocked)

I just came back from Hannover And If I had money, I would return for Cebit (my mate lives just a short walk from Messe ). Unfortunately, it's nearly impossible to get a job here it seems... otherwise I could've asked for a week off, after having earned enough for a trip damn Swedish kids:p

I suggest that someone makes a new account, "xxxx's grandma", posts some nice 3D benches with HW he's already used for some submissions, and then posts a picture of his 90 year old grandma sitting on a chair in a nursing home, with a 980x on a plate, next to a cookie. The way alot of people think of this situation, we can't touch that person who stands behind those two accounts. "No proof". Gah. Not saying that Dejo's daughter doesn't exist, but we have to draw the line somewhere. If we had perfect tools to discover result/HW sharing, then this thread wouldn't have existed in the first place - but we don't. We just have to draw the line at some point ("suspicious enough"). By looking at those screenshots posted earlier it's not hard to imagine some flashing, red lights in the mod section, no? We can always move the line a bit, so that it's MUCH easier to set up multiple accounts and get away with it - which will also make life easier for Dejo and his daughter. I don't know what you guys prefer, some LIGHT restrictions one ONE account, or a gigantic loophole where teams can get flooded by "relatives". That's why massman did a very nice job by suggesting other solutions, so we wouldn't have this ridiculous discussion at all. It seems as people don't realize that you can't have a perfect solution to all problems, most of the times we have to find the better one.

it may not be against the rules, but the loophole is gigantic - as said before. We can't ban an account if "both users" say they're living under the same roof, and therefore share alot of HW and OS installs - there's no way we can prove that this other person is not real, or using the same CPUs/GPUs => multiple accounts/HW sharing FTW.

I can confirm that the submission won't be blocked. maybe i'll "ok" it right away:D But if you've got the screenie, feel free to reupload it:)

I've seen this error on legit submissions as well. Too bad it's impossible to know. personally I'd OK the score, as it doesn't exactly look bugged low, but if a decision has been made I won't undo it unless the owner complains.

Not meaning to trashtalk in this thread, but where are the 01 numbers?

happy Birthday. I hear Massman is finally old enough to drink now:D

If old scores without proof look unreasonable, they will be removed.

Looks like this one has been moved already:)

I can't recall ANY of these submissions being validated. Ever. Old submissions or new. it would be easier if the program would just stop running after the error happened, but apart from that it's just as when a superpi run stops at loop 14. It makes no sense to call the result valid, as the computation is different from the results you compare it to.

Any particular board, or all of them? No-one compared EVGA to the lot?