Strunkenbold

Right, it is more correct to report those bugs. Will see how it can be done. But I just see so many bugs, I dont where to start and when I should have time for this.

The more I dig into this the more I understand that there is a big difference between what CPU-Z is detecting and all other Internet databases. If I could I would fix this in CPU-Z but Im not sure if this is even technical possible. Im totally with you having our db correct like CPU World or Intel site would make everything a lot easier and more correct. So now just 4 fun a common issue in our db: User submits a result with some old CPU, CPU-Z tells him: Pentium M with 1Ghz and socket 479 When he try to submit using the autocompletion he gets nothing because there is no Pentium M 1000Mhz or Pentium M 1,0Ghz. So he ends up searching the right one in the db and when he is lucky enough he might select Pentium M ULV 1,0Ghz as the correct one. This might be successful but is this really necessary? What do we gain by this? Detection of CPU-Z is really that bad as I dont even detects a Pentium M 723 like here. So it gets in the wrong category. So maybe we could do every user a favor and make our name styling a little more CPU-Z friendly even if its not 100% correct. So in this both cases I would propose: Pentium M 1,0Ghz ULV (Banias) Pentium M 1,0Ghz ULV (723) just my 2c

Not possible. Your score with 2211 points is currently your closest result to the target, or it is even already the target score.

Already received the informations?

Its because the T9400 has accidentally the wrong socket in DB, opened a new ticket for this. Should be correct if Antinomy agrees to my changes and Massman update this list. Edit: But its still funny to see how Hipro5 humiliates those Core2 Cpus...

Move to BGA479: Celeron M 205 Celeron M 215 Celeron M 310 Celeron M 318 Celeron M 353 Celeron M 373 Celeron M 800Mhz Core 2 Duo U2400 Core 2 Duo U2500 Core L2300 (1.5Ghz) Core L2400 (1.66Ghz) Core L2500 (1.83Ghz) Pentium M 758 Pentium M 753 Pentium M 738 Pentium M 723 Pentium M 733 Pentium M 733J Pentium M 773 Pentium M 778 Pentium M LV 1.1 GHz Pentium M LV 1.2 GHz Pentium M LV 1.3 GHz (718) Pentium M ULV 0.9 GHz Pentium M ULV 1.0 GHz Pentium M ULV 1.1 GHz (713) Move to BGA953: Core 2 Duo SU2300 Core 2 Duo SU2700 Core 2 U3300 (1.2Ghz) Core 2 U3500 (1.4Ghz) Core 2 U9300 (1.2Ghz) Core 2 U9400 (1.4Ghz) Core Solo U1300 (1.06Ghz) Core Solo U1400 (1.2Ghz) Core Solo U1500 (1.33Ghz) Move to Socket P: Core 2 T9400 (2.53Ghz) Rename: Pentium M LV 1.1 Ghz to Pentium M 1.1Ghz LV Pentium M LV 1.2 Ghz to Pentium M 1.2Ghz LV Pentium M LV 1.3 GHz (718) to Pentium M 1.3Ghz LV (718) Pentium M ULV 0.9 Ghz to Pentium M 900Mhz ULV Pentium M ULV 1.0 Ghz to Pentium M 1.0Ghz ULV Pentium M ULV 1.1 GHz (713) to Pentium M 1.1Ghz ULV (ULV) Core 2 U3300 (1.2Ghz) to Core 2 Solo SU3300 (1.2Ghz) Core 2 U3500 (1.4Ghz) to Core 2 Solo SU3500 (1.4Ghz) Core 2 U9300 (1.2Ghz) to Core 2 Duo SU9300 (1.2Ghz) Core 2 U9400 (1.4Ghz) to Core 2 Duo SU9400 (1.4Ghz) Core L2300 (1.5Ghz) to Core Duo L2300 (1.5Ghz) Core L2400 (1.66Ghz) to Core Duo L2400 (1.66Ghz) Core L2500 (1.83Ghz) to Core Duo L2500 (1.83Ghz) Celeron M 350 to Celeron M 350 B1 Celeron M 350J to Celeron M 350J B1 Celeron M 360 to Celeron M 360 B1 Celeron M 360J to Celeron M 360J C0 Pentium M 715 to Pentium M 715 B1 Pentium M 715A to Pentium M 715A C0 Pentium M 725 to Pentium M 725 B1 Pentium M 725A to Pentium M 725A C0 Pentium M 735 to Pentium M 735 B1 Pentium M 735A to Pentium M 735A C0 Merge: Pentium M 745 and Pentium M 745A As its seems that there are also C0 revisions of M 745, there is no way to distinguish. Hope everything is right...

Das erklärt dann natürlich einiges.... Sah jedenfalls nach einer spaßigen Competition aus konnte leider nicht mitmachen.

Ihr müsst euch halt ein Bett teilen aber so kommt wenigstens richtige Camp Atmosphäre auf. Und hättest du einfach eine CPU-Z vali gemacht wäre das gar nicht passiert.

Well its over. One hour after dead time I also got 18 points. But too late. Congrats to Coldest for winning this.

That was the first thing I thought of but running a second program side by side is crashing the benchmark process. You really have to extremely reduce bclk and tweak to steal some cpu cycles with programs without letting XTU crash. Taking into account that one run takes up to one hour this was for sure the hardest stage of the competition.

Its because the selection for pifast is wrong: cpu_model.socket_id=269 or cpu_model.socket_id=120

XTU 5.1 is very picky, it seems 4.4 is far better. I assumed its not allowed, because its outdated but apparently it works. Just finished a Run with 87,75% clock modulation which results in 18 Points. grrr

Damn that last one drove me crazy. Managed to get 81.77Mhz but result didnt lower. Relaxing Ram timings resulted in "benchmarking process has failed" also the same with playing with Clock Modulation. What was we supposed to hit in last one?

Yup agree here, should be really only best user result. A Top10 is to short for multiple entrys by same user. BTW, did anyone looked at first spot in 3dm01? 2700k misinterpreted as Athlon 2700+!

If I could, I would vote for: Slot A Socket 754 Socket 939 Socket 370 Socket 478 Socket 479 (someone need to clean database first, because of Core2 CPUs which dont belong there) Socket 423 can go

Sensationell 5,3Ghz mit den niedlichen Lüfter!

Thx very much Massman! People will like this.

Very nice bclk

Im not disappointed of your efforts, its quite the opposite! I expected nothing, not a single response but now we have 2 threads full with people showing interest and this led to a new competition series, not bad huh? If anything, Im a little bit disappointed that _I_ cant properly explain you why such a ranking system could be great fun. (and I failed very hard, as it seems that nobody understood my first post at all) I still thinking main advantages would be: -cheap parts -_almost_ no binning requirement -relies mostly on benchers skill I just thought that this would be an interesting concept, it could be a tribute to the golden days of OC and maybe we even can involve people left the scene already some time ago to return. Maybe they just need a platform again to show their skills. But I have to admit, I didnt thought it would be so hard to implement this. I totally understand that this have to wait till Devs have some room to breath. As an alternative, wouldnt it be possible to create such HOF rankings like you did with the 5Ghz sp32m challenge? So saying, creating a thread for i.e. Socket 462 and include Top20 for CPU-Z, SP1m, SP32m, Pifast, wp32, wp1024 and so on?

If I red it correctly, Massman is just about to introduce yet another league anyway to motivate some youngster. This raises the question what does Massman to motivate people which are already some years in the game and dont want to become the master of CPU binning for Xtreme League or like to bench 100 different Socket 775 CPUs on LN2 for TopSpot in Hardware League. Also we should consider that we already have a separated ranking in the bot, its called "Mobile Masters League". Is it over the top to create another menu point and accumulating some specific points together and call it Old School League? But I try to be realistic, I was surprised about the response this threat got, I mean we have monthly competitions back which is very cool! But also knew that Massman wont spend coding time for old stuff which no sponsor is interesting in.

So this polymer caps in side to the smd ones really helped that much? Great record btw, I guess this will last some time.

I guess you bite into your desk at this moment.

Sorry but thats exactly the point, benching oldschool stuff doesnt do anything in hardware league. You can bench Socket 7 Systems forever, do crazy mods spend LN2 on it and get 100 pts in the end. Because benching that stuff awards you with 2pts in the best case, this is dedication for sure but to gain ground you have to own a good 4770k/4790k go to ebay and buy Geforce 8800 or similar. Last time I got a Radeon 5850 for cheap. I didnt even had to fire up my soldering iron or install a waterblock still I got 20pts. Dont get me wrong, I know these points are cheap to get and will lower with every new CPU generation. To stay on top you have to do some really good CPU, mainly Socket 775 results but benching all the real old stuff is mainly for the team. So thats why I had the idea for a separate ranking system. Introducing a Hall of Fame page for each socket would be the first step. You could award top positions with badges, which appear on users profile page but real competition comes with points!

Hey its very cool that this really happened. I would like to thank anybody involved with this. Although I actually didnt want to re establish the monthly competition series I appreciate the work made by the mods. My intention was to add a separate oldschool ranking to the bot. Seems I either didnt explained it correctly or its just unrealistic because of the huge coding work. When I could add something to Round 1 of Competition: We should maybe think about lift the limitations a bit to reach people participate because of "show & shine" factor. Not everyone has exactly the hardware needed lying around or wants to ebaying. So it may be a good idea to allow Geforce 1-4 also and sockets appeared prior S478 and S939. They wont have big chances to win but atleast they can participate. Superpi 1m 2801Mhz Cap is a nice idea but it could also be interesting when you add a full out stage. Socket 478 standing no chance against S939...

Can confirm this...