knopflerbruce

So far I haven't seen this score being compared to any specific ones. If it - for an unknown reason - is 50% higher than other scores that can be used for a comparison, then it needs some explaining.

I mean a similar configuration.

Are there no similar scores this one can be compared to?

I'll bump this. Any particular things you guys have in mind - high priority stuff? AFAIK we updated a few things not so long ago.

This is true, but perhaps it's a good idea to make a newspost about these tweaks that we already know will be forbidden - instead of waiting for ALL the clarifications? It puts the benchers in a weird situation where nobody knows what we accept or not.

I don't know who have told you to do what (definatly not me), but there are no restrictions on CPUZ version written in the AM3 rules - effectively that means what I said: use whatever version that correctly displays frequencies and so on. That's the only reason we want CPUZ tabs to be included in the first place. Same for the general rules, no version requirements. And if someone pointed you to the CPUZ benchmark rules, well... that's a mistake. CPUZ benchmark rules have NOTHING to do with AM3, or any other benchmark than CPUZ:)

It would be extremely harsh to block a sub because an old CPUZ version was used for CPU speed etc. The important part is that it is recent enough to display the information correctly. The last link you used is also for CPUZ subs only, not 3DMark. The 3DMarkxx pages have no limitation on CPUZ version I think. I wouldn't mind having mem tab mandatory for 3D, but that's not the case today. Perhaps time to discuss it, i think most people add it anyway.

Got email from Gigabyte, boards are shipped wohooo!

Yeah, it's rather unfortunate that this particular chip is the best one Got a 3850 chip that show the correct model.

http://www.xtremesystems.org/forums/showthread.php?64370-Fx&p=889817&viewfull=1#post889817 - yep, I can post a picture later of the CPU. Not so easy to find it among so many others ^^

Heard nothing either.

Maybe Ryba and Gigabyte can compete... who is the slowest to deliver promised parts wins free advertizing space here !

512 it is then. No k:D

I wouldn't change sector size if i were you - leave it at the default value (512k?)

There's a bug when using the prepopulate function when submitting scores. If you have, say, a few singlecore subs you want to submit, and you use a dual/tri/quad (etc)-core, the score will pop up as a 2x/3x/4x score even if you have changed to the singlecore model. However, if you do it the other way around, use a singlecore sub as "filler" and then change CPU model to a quad, then it's listed as 4x.

Try to calculate the total benchmark running time

The general problem with RST is that it has been used for a long time. It's kinda the same as blocking LOD NOW, although on a smaller scale. it's impossible to clear the bot for RST based results, banning it is perhaps the right solution, but it takes too much work to go throught he whole 05 database looking for those results.

I think both use ramcaching. They're pretty much equal apps if you don't mess with the sector size or what is was causing bugged scores.

Added to the thread in the classified section

OK, so we should add 512k HDD sector size to the rules?

The problem about RST is that it uses some sort of ramcache, no?

Agree:D

Make a newspost. Then we can punish those who do not care.

it's correct that the common practise was to ignore these rules. Perhaps it's our fault that we didn't investigate this earlier, but it's any bencher's responsibility to make sure that the tweaks he uses are allowed - accepted practise or not. If you find a tweak that gives a huge boost it's common sense to verify that you're still within the rules. Everyone who used them were free to share the tweaks in public, or tell us that they were not legit as pr RULES, but they chose not to. That part has nothing to do with HWBot itself, only those who chose to use those tweaks. That's the mess we're clearing up now, and I'm happy to see that people understand that what was accepted practise for a while was wrong and that we have to block a bunch of valuable scores.

There are no new rules. We go after those who replace files and in other ways change the calculations in inappropriate ways, or creating bugged results. this has been illegal since the very beginning