knopflerbruce

Yeah, any other apps would be additional, not replacements for CPUZ;)

What would be nice to have is a program that would work this way: rank based on number of pi digits computed in (fixed time) seconds. That one would last "forever", since the score would increse with time, not approach zero, as we see with superpi 1m, and especially wprime32m. No idea if that's doable with this app, though...

One bench that needs to be added is one that lasts 1h+ for current CPUs. We had it in the past, but not anymore:rolleyes:

Yeah, but only 4 of top 20 haven't reached the 300. Perhaps it will affect the lower parts of the rank more, but that one is still messed up by the way too low limit of HW points so it's not like it ruins anything:p Still, the 400 point limit was just an exampl. Raising the VERY popular ones to 20 or 25 won't change too much, I think. But this is Intel-business, so I'l lleave it up to you. Doesn't affect my scores one single bit;) Unless xeon 3120 will become insanely popular:p

Well, those 2 rankings are nearly independant of each other - the 300 limit takes care of that. Giving like 400 points for the top Q6600 wPrime 32m score still doesnt affect the global points ranking:)

It's a damn close race between E66 and Q66.

Yeah, the last one has crossed my mind, too. I think E6600 (and the others in the same "league" submission-wise) deserve a few points more, perhaps 20 or 25.

The part number in the picture in the link says it's a 2.6ghz B2 chip (there is a 26 early in the PN, and it ends with BGD which is proof that it's a 9900). Yeah, because of some validation bug... it did link to the correct one when it was created, but then it changed... i have too many validations for CPUZ i think:p

Sorry for going OT, but do you mind adding this one, too: http://www.hwbot.org/forum/helpcenter.php?do=ticket&tid=208 ? It's been in the queue since January...

If it's going to be added, then the slowest test should take about 2-3 min. at least - since it's multithreaded (no need for another 2 sec benchmark...), and the slowest one 1h++ There are too few benchmarks here that need some time to finish - even superpi32m is over almost before you've started it:(

Just curious... what CPUs are we talking about here?

A permanent fix will be difficult, but you can always PM the owner of the other score, and ask him to delete, and reupload it. Then I think your score will be first.

I always wonder... why can't these cards just be added? Then we have no arguing about unfair scores like this, and more people will bench their cards in more categories. Why was 2900pro/xt separated, btw? Same cards:p 7900gto vs gtx?;)9800gtx vs 9800gtx+? Just wonder what's so different there from this discussion.

You need GPUZ after 1.1.2008, or an ORB link. If they're not following the rules, report them.

That one should either be blocked, or moved to 512bit...

Note that you can share memory, PSU etc... just not GPUs (for 3D) and CPUs (for 2D). for exampl, I think you can use the same E8600 (or whatever) for 3dmark01, but not the same video card. However, for superpi you can use the same video card, but not the same CPU.

I'd say it's OK to post it, as there's no doubt that there is a bug with PII and CPUZ rejections...

This is a well-known error with that chip. Should be OK to upload the scores to the bot despite the fact that it's rejected.

Alot of people need to sell their CPUs/GPUs to be able to afford to hottest gear, not allowing selling would make it alot harder to bench. Buying a proven overclocker is not wrong, borrowing it for a session is.

Not if the S/N is stored on some chip that can't be flashed;) It's not realistic, I know, but it would solve the problem.

You can't really start being suspicious before the results are uploaded lol:p

The question is not what the GPU clocks are, it's who submitted first;) The oldest submission should be above the newer one.

I've seen that before. I'm glad the other version is so easy to spot;) More scores that need to be moved/deleted: http://www.hwbot.org/result.do?resultId=500005 - DFI never made a 940 mobo. And the clock is WAY above what these chips can do... http://www.hwbot.org/result.do?resultId=444773 - Says DFI nf4 Ultra-D, which is socket 939 - not 940...

http://www.hwbot.org/result.do?resultId=785877 It's like 35% too fast for the clocks:p No doubt it's not the correct version. No CPUZ either, just some other program.

Agreed.