knopflerbruce

I wonder why there are so few AMD samples out there, and so many Intel's:D If Intel cared about this, you wouldn't find them on ebay, as the auctions would be removed. Intel knows about ebay, and if they choose not to report the auctions they indirectly say that it's no big deal to them.

If we can't see what CPU it is, then it can't be accepted, unfortuantely. However, you can always try to start CPUZ before running 1m. Maybe the efficiency will drop a little bit (set CPUZ to low priority to minimize the effect), but at least you'll get a good result.

I'll try on socket F, 939, 754 and AM2 at some point, but I have an exam I must do before I can create enough space here for another setup:D

One solution is to jsut reupload it, and add some random crap in the description field (to avoid duplicate message, (maybe this is fixed in rev 3.5?) )

Let's say Fiat Uno then:D

If you can store 500 results in each benchmark with the free version, then it could be OK to always ask for a link. If not, then it's pretty useless - as they will be deleted fast. And nothing is more annoying than verification links that don't work

Damn, I need a REX so I can bench my GPUs in 01 and 03, so I can keep up with you freaks:D Gaining boints from superpi 32m and wprime 1024m on socket 754, 939 and am2 takes forever:o

A little late, but congrats anyway:p

Ticket ID: 840 Priority: Low TwinMOS is missing from the manufacturers list. Also, I noticed that the VX and VX2 series of OCZ memory aren\'t listed as well.

http://hwbot.org/hardware/processor/athlon_64_4600_x2_brisbane?tab=2drankings#cpu-z Bug: CPUZ seems to be divided by # of sockets, which it shouldn't as CPUZ is a singlethreaded benchmark. It may also be an idea to remove the multi-socket and unlocking options for the CPUs that don't have these features (example: this chip doesn't run in dual socket boards and can't be unlocked), to remove the "misplaced" submissions where people count cores and not sockets.

Haha, that's EXTREME - backup for 0.5 points:D

The paypal text makes the "most popular" picture dive into the ranks when I visit this page: http://hwbot.org/browseHardwareProcessors.do?cpuSubFamilyId=5 Running FF 1280x1024

I guess this is just very temporary, then. I bet one of them will change by 0.1 boint soon. Still, is there a possible solution?

I bet it's because they're equal.

http://hwbot.org/forum/showthread.php?t=6267 - doesn't work here, and I had the same problem myself. Seems to be the recalculation that fails.

They lost the boints because of the edits, so you you should reupload them in the correct ranking.

The editing itself works, but the submissions lose all the boints after the edit (says it's not the best, even if it is).

Then you should send a PM to all mods so they don't change existing scores, as I had a typo in one of my submissions and I had to reupload to make it work again after a crew member "fixed" it.

THis one is from 2007, so other versions are OK. The 1.55 rule was applied 1 Jan 2008, and of course older submissions are still valid.

Bug: http://hwbot.org/hardware/processor/athlon_64_fx_55_clawhammer?tab=2drankings#superpi The score that's currently on top shouldn't be there. Also, in alot of charts there are many results that are not linked to profiles, and that don't receive boints. Why are those visible? Or maybe a better question: why aren't they rewarded with boints when they are ranked?

1.55 became the only allowed version on 01.01.08, before that it was more like "anything goes", apart from a cheatable version (that was 1.4x if I remember correctly, not 1.53, so this version was allowed at the time)

Numpad error I'll edit it later, when the edit bug is fixed (or someone posts a score between 46.73 and 49.73, atm it's not stealing any boints from anyone).

I'll leave this decision to the Nehalem experts in the crew, but if Kurumi reads this thread he should know that a superpi run at 5.4-ish (with a time that indicates 5.4GHz) is good proof that this bclk is real.

You probably need a 980 to be able to get the top spots, but there are so many that receive globals atm, that you'll get plenty even if you're not running the $1000 parts. 980x will hold quite a few records, because it's the fastest chip around. That's not so strange. PS: there are hardware boints to fight for as well, you know... PPS: E8600 still rocks 01 and 03...

The edit bug is still there... Some guy posted an Opteron 144 90nm result in the 130nm category, and I edited it. Now the guy doesn't get any boints for the submission, as it's "not his best in the category" anymore.