Strunkenbold

huh that score is really some years old now. Cant say which voltage was exactly responsible. You probably just have to try and find out. Just make sure you change thermal paste on nb and always have some fan blowing at the cooler.

There seems to be some db issues. 1) Submissions with i3 3110m get constantly linked to i3 330m. This especially happens with XTU an HWBot Prime subs. 2) There is something wrong with the speed of the P3 Celeron 566Mhz, as the db says its default clock is 533Mhz but obviously should be 566Mhz.

Is there any reason that I need a screenshot for #1 and #2 while for the memory competitions I just need the verification link?

AFAIK there are no boards with support for SLI/CF. Otherwise, making this fullout would make this stage cpu limited like the 3DM01 K6 stage. Very interesting though. Cheap platform also. 60-70€ for cpu+board.

thx!

Ticket ID: 1907 Priority: Low #1 Wrong socket for:\r\nAthlon 5150\r\nAthlon 5350\r\nSempron 2650\r\nSempron 3850\r\nCurrently these are BGA FT1 should be Socket AM1.\r\n#2 Athlon XP-M 2100+ Barton currently Slot 1, should be Socket A\r\n#3 Rename BGA812 to Socket ASB2\r\n#4 Move following cpus to Socket ASB2:\r\nV 105\r\nAthlon II Neo K325\r\nTurion II Neo K625\r\nTurion II Neo K685

Also just a PCI-E card. No GPU-Z shot available but its clearly visible from the system pic...

NP. We all just started somewhere...

That was actually exactly the same what I was thinking when I saw this...

This is not a PCI card... In case of a real pci card, GPU-Z would say @x1 1.1, not @x16 1.1.

Bugged GT2, you better try to reduce your clocks and get a valid run.

+1 But also keep in mind that, when we only accept default thread count, probably 40-50% of results have to get blocked. And than we have to think about if it makes sense to keep this bench.

Actually 12k results are way too much to get moderated by 2-3 persons. This will take ages. We probably should consider making a clean cut here. There are more up2date benchmarks out there dunno what is so special about UCbench that we have to keep it.

Please imagine people who want to submit and dont know our db by heart. They look up for what CPU-Z tells them. So take for our example "Sempron 2800+ Palermo". They enter Sempron 2800+ in the submit form. Two results appear, both Athlon XP one's- seems that this arent the right one's. Next enter just Sempron, damn over 10 results, ok than 2800+, damn also over 10 results. They end up looking in db for them self. This extra "64" in the names has no advantage. For what should it be good for? We already have a name scheming for cases like this. Core name in brackets is enough. Also implying not existing cpu extensions is simply plain wrong and just another reason for removing "64" in those names...

Whats the exact case for single cores? There are obviously bugged scores as well using just 1 and 64 threads. And 64 would be even. Or are these results ok?

Currently there are just two more results when I enter "Sempron 2800+" in the searchbar. Its quite the opposite when I enter "64 2800+", I got 10+ results. Really, your doing the user no favor by putting a synthetic "64" in the middle of a official product name. Actually this needs to be removed for any Sempron, not just the mobile ones.

Ok I think I got the idea behind this. But when I enter in the quick search what CPU-Z gives me I will never find those cpus in the db. As Im typing in Sempron than i.e. 2800+ I always get those Athlon XP one's. Isnt it enough that we distinguish them by core name?

Ticket ID: 1895 Priority: Low Hi,\r\nall socket 754 AMD Mobile Sempron 64 cpu\'s, in the db are wrong named. \r\nThey have no 64bit extension, thus are simply just \"Mobile Sempron\"\r\nSee also here: http://www.cpu-world.com/CPUs/K8/TYPE-Mobile%20Sempron.html

LF would mean Roma. You can check yourself using OCInside interactive rating & product ID guide.

nice!

So that means with higher thread count, pwds calculated need to get higher? (I actually dont understand the technical background to this "rule", nor do I see any submission fulfilling this, but it would be nice to see some clarification about this to see what a bugged run actually is)

Thats true, but this is same situation like we had with wprime. Old versions had a major performance advantage. You cant beat those scores even with better clocks. Dont know what would be a fair solution but as maxxmem doesnt receive any points anyway I would say let them disappear...

IMO db should be cleared with those old versions. If I remember correctly, there was some discussion about incorrect/cheated results.

Thx for info! I wonder what this jumper did. Guess if someone would figure out, other boards could be modded as well.

Cool setup! Did you had to mod anything? As on my Cuv4x-d, two Celerons wont boot...