knopflerbruce

Massman's arguments were not about the version, but about how important this tab actually is. In fact, this can be any dual core conroe/wolfdale - we can't tell. If his suicide shot is at 5133, I would think the cpuz speed was a bit lower than that. If you can allow the first spot in a category that has 1174 results to be without a CPUZ tab you can ignore that rule for top 20 global submissions as well. Will that ever happen? I hope not.

http://hwbot.org/forum/showthread.php?t=3859 Consistency => block. I also had a few blocked scores a few months ago because of a missing memory tab (could only open one cpuz window...), and if thats a reason to block then this definately is. Plus, this score is #1. IMO #1 scores should ALWAYS have proper screenshots. That's the one most people look at, and alot of folks will wonder why there is no proof of CPUZ type there... The CPUZ window is without a doubt the most important validation part, except for the score itself of course, and if that's not mandatory then you can't really moderate other scores at all:)

You can buy stuff from other countries... there are always some ebay sellers that ship worldwide;)

...so if the 80nm 7600gt had been called 7600gt+ it would've been separated from the rest? I thought this was about difference in performance/overclockability and not whatever name nvidia put on the sticker on the card.

Different process is a good reason for a split IMO. It's like comparing Conroe to Wolfdale if they have the same cache, and we don't do that... But it'll take forever to check all the 7600gt scores:p (however, the difference in OC capability isn't very big, seems even 90nm can hit insane core speed under cold. On the other hand... the same is the case with 9800gtx and 9800gtx+, and they still got split. Consistency FTW!)

It says "recent" version, which doesnt mean the newEST, just that it has to be fairly new.

Lmao:D

...but if OP actually does what I say and posts his results in some category, will his submissions be removed because we don't have - or want to create - a category that fits with his card? Or is 9600gt ddr2 a decent category to use in this case?

...so 9600gt ddr2 it is:) If there were a million different 9 series cards i'd agree with you, but atm there are not too many - this one isn't even close to fit in any category. I'm sure that if more "weird" cards show up then at some point there will be no need for more rankings as the specs are always pretty close to some other card that's already in the database. OP SHOULD be allowed to put this card in a category that's fair for him and the competitiors in that category. If we can't add hardware because it's not "common enough" then HWBot is moving in a very bad direction... Plus: http://hwbot.org/forum/showpost.php?p=28268&postcount=37 This is not a "complete VGA database" if you don't include cards that are quite different from the reference model

There should be a category where he can post scores with this card, a category that fits the specs and performance of the GPU. 9600gt ddr2 may fit performance wise. 9600gso is out of the question, it's way faster because of the 96 shaders. Same with the regular ddr3 9600gt category, because of the shaders. The DDR2 memory may compensate a little for the missing shaders. On the other hand: posting a score with a card that has the wrong memory type as well as the wrong amount of shaders isn't exactly optimal. However: if creating a new category is not an option that's the only thing you can do

Ticket ID: 626 Priority: Low http://hwbot.org/result.do?resultId=910806\r\n\r\nSeems like the ddr3 memory card gives a very nice boost compared to ddr2, so a new category should be made.

Damn school steals all my time...... it's my job to move the scores.

Do the scores reflect it being a 9800gt? Or maybe they fit perfectly in together with the 9400gt's?

I can see the "need" for backups... like, it's always an advantage to be in the last pair during a skating competition, as you know all the other track times and then know what times to beat. But benching takes alot more than 5 minutes, so if you post your best score during the last hour, there's little to no chance anyne can setup a new rig and squeeze the absolute max out of it and make the deadline. Which is the number one reason for me to say it's not a suggestion that makes any positive differences. ...and on the other hand, is it common to not post the best scores unless you really have to? Like, you have the top score until the last few moments, but you have a backup thats higher. Who would you NOT want to post that one near the deadline? Sounds incredibly stupid to me:D At least in a competition like this where you don't get global cups for your work, only HW cups at best.

Dannnng, now I'm 5th:p Well done, bud!

We're discussing this in the crew only section as well;) Just to let you know we're working on it...

Superpi needs a memory tab:p

Unfortunately I can't make rules alone:p I pointed out this a long time ago, maybe something will be done about it now that alot of users react:) A screenshot isn't that much to ask. Even memory tab is unnecessary, as it's nearly a memory independant benchmark. Maybe it's different for really old CPUs, but not for K8 - and I doubt for anything newer than that. CPUZ + wprime screenie should be enough here.

I agre, screenshot should be required for wprime as well - we all know the direct submit part of the program doesn't work like it should in alot of cases.

Hmmm, I don't know what to say about that one. The same optimization halps in some games, but perhaps not enough of them....?

THe most obvious way is to go to "overclocking", and then click on world records. Which gives you this page: http://hwbot.org/rankings/worldrecords - no singlecards there. Edit: the list on the right side of the front page gives the page I'm looking for, that's correct

There is a link in one of the pifast (?) submissions to a thread where two CPUs are pictured.

OT: the singlecard HOF is quite hard to find I think... I had the same problem, I could only find multi-card HOF when I wanted to look at some single-GPU scores.

Then this case is pretty clear. Why wasnt it included in the first place, btw? The program had to be opened to be able to make a validation file...

I would say that if the memory frequency in the cpuz mem tab and the one in the validation link match, as well as core temp displaying the correct cpu, and the score is normal for these clocks, then it should be possible to OK it under the "common sense" guideline. PS: the format of the screenshot is awful, why did you make it so wide? I can barely read it on my 1280x1024 monitor...