http://membres.lycos.fr/cpulloverclo...matique/4G.jpg
http://membres.lycos.fr/cpulloverclo...-233.86pai.jpg
Taken from digital-info.net forums, quite an accomplishment
Printable View
http://membres.lycos.fr/cpulloverclo...matique/4G.jpg
http://membres.lycos.fr/cpulloverclo...-233.86pai.jpg
Taken from digital-info.net forums, quite an accomplishment
wow.... what voltage was that at?
I think its fake, could be wrong just from a glance doesnt seem all to real. Start super pi @ 5:43 and finish @ 4:53?? Thats a long 19 seconds. That and i've never seen a dothan clocked that high. :toast: Would be cool if im wrong.
Thats insane. Dothan has matched FX in both PI and 3Dmark (actually beaten it)and in mhz. Cant wait for an 18sec run.
Cant really say I believe it myself...
Cpulloverclock posted it earlier today... I think it's OK, but I don't like that he did not use msec version :( Now, we don't even know if it's patched or not, although cpull said it's un-patched...
very nice if real
Fake... must be a fake.
It just doesn't make sense to clock CPU with that design so high..
Remember that Dothan has very low latency L2, high IPC etc..
So cooling with what ever wont take it to 4Ghz.
But i guess that it would be ok overclock if it's real.. :D
That doesn't make sense, if macci breaks 20 with 36xx... and that guy has 4ghz..
Doesn't make sense. It seems to be a fake to me...
http://u-san.net/c-board/file/14-278.65-CPUMark-320.JPG officcial 3.9Ghz dothan..
So why not 4Ghz ?
But im still suspicious, especially if the difference to other clockers is high...
What are best clocks for Dothan that multiple users have achieved ? (within 1-2%).
Of course it's real you guys, lol...
It's the new 780, clocked at 2.26gHz... on LN2, the Japanese Aces have been dancing around 3800-3900superPi stable for MONTHS now, with 750's and 760's.
Yeah, it's real, and yeah, it's fast. And in Spring of next year, 2 of those clocked at 667fsb will be on a YONAH CPU...
Not only is Dothan "faster" than FX-5X clock for clock, but it'll even prove to clock higher at the end of the day, also...
C
Where can I trade my Neo2 and Winny for a Dothan w/ CT-479 and Asus P4C800-E?
The "only" thing that will save AMD is the motherboard "support" factor. When AMR boards are available, users will be forced to use AMD Athlon 64 CPU's... and there's no SLI possible on the P4C series with the CT-479 adapter...
So, bottom line, Dothan IS the KING of CPU's...
A KING with no "KINGDOM", if you may.
someone needs to put out a 479 sli board, methinks the first one to do so will reap a great amount of rewards. Or wait and make it an AMR board.
But, why is it so slow...Quote:
Originally Posted by charlie
sif its a fake...
dothan pwnage.
awsome, he could have upped more the fsb 230 isnt that fast.
it doesnt matter if it is patched (i think) dothan doesnt support sse3
must have a yonah!!!
if its real.... drool... but somehow i dont think it is.. id like to see a cpuz verified shot.
Dunno, but TAM and some of the others are at 18s and 3850mHz, VERIFIED, why is 4gHz with a new CPU so hard to believe???Quote:
Originally Posted by urbanfox74
AMD Kool-Aid drinkers.......
Quote:
Originally Posted by charlie
All I'm saying is that it just doesn't seem to add up, he is 350+mhz higher than macci yet he is within the same second...
I think that's a rather innapropriate thing to say, you don't know me or my attitude towards different companys. I would love to go out and buy a dothan setup but I don't have the $$, so please jsut because I don't believe the validity of some SS don't label me a fanboy...
4 Ghz in a dothan in insane...
But I don't understand why it only gets 19s. I would expect about 17 highs-18 lows ...
4000 Mhz are a lof of Mhz of power. Maybe generic memos :p: ?¿?¿?¿ , I don't think so in someone with a 780
Why would a famous japanese overclocker kaz-n fake his screenshot. Also 4 GHz with 780 and LN2 is very much achievable. Don't you guys know anything about japanese culture? How many fakes have you seen from japanese well known overclockers? I haven't seen any.
I think he was able to run Spi1M @ 3975 MHz (when time was 4:53). Then he tried to run it at higher speeds but propably managed only get a screenshot of 4 GHz.
Well, let's clarify a few things...
His superPi run is at 3975mHz, 17 x 233, 1:1... probably slack timings.
The 4000mHz is just a screenshot.
These things are VERY sensitive to memory bandwidth, and 233 isn't much, especially at slack timings.
And HELL YEAH, Sampsa is right, these guys don't B.S.
I BET that 16 x 250, 1:1 2-2-2 with FULL Win2k3 Server tweaks would be 17s.
C
Maybe it's that he is running the memory only at 233 2'5-3-3 :p:
Thanks so much Charlie, Sampsa & other XS members for your trust on members of Team Japan :toast:
Also, cpulloverclock to report initially our activities in here in XS with many tks ;)
BTW, there are some certain rules/restrictions although unspoken but which are well respected by Team Japan members for running SuperPI as follows:
1. You may disable some OS services,
2. You may adjust size of pagefile,
3. You may use memory tweaks such as setting by maxmem or A64Tweaker,
4. You may adjust L2 Cache size in registry,
5. You may set Priority in Processes at your option,
6. How you format & partition HDD is up to your option.
It is quite arguable what tweaks would be acceptable but these are the rules atm after a quite bitter arguments or battles you might say at Team Japan when one of extreme OCer came up with his tweaks to end up 0(ZERO) second on SuperPI 1M in the past.
But, trust or not trust is all up to your discreation ;)
Cheers
Come on guys, it's 100% real. When a Japanese well known overclocker shows his results I don't doubt the results at all. Kaz have changed his superpi results like once every week almost and have been around 4ghz so I'm not surprised he made it.
Congrats Kaz and Team Japan.
Yup, Charlie, you are right when you read his explanation but unfortunately in Japanese....Quote:
Originally Posted by charlie
He said he used 2 x 256MB memory and would try with 2 x 512MB next time..
Also, he doubts he can get 18s...
Hope to see his results soon :)
Hi Fred :wave:Quote:
Originally Posted by PRTeamJapan
285*14 1:1 2-2-2-5 :D
ln2 on the NB, cpu and memories
-170C
vcore 1.7v
vdimm 3.8-3.9v
vdd 2.1v
@PRTeamJapan
Any chance that Kaz-n can start using SuperPi mod1.4 w/ msec. I looked at his website and pretty much he hasn't used it yet.
Yup, already asked him to do so--Quote:
Originally Posted by bachus_anonym
Oh well, we shall see as these guys are such stubborn, living in their own OC World you might say :D
BTW, mod1.4 is quite weird-- so much variances each time I run:confused:
I got 24.203s @3433GHz
a better 24.000s @3423GHz
May be mili second portion is not so important :D
yes, I got 25.2 at 3215 with it and 25.2 always at 3250 :stick:Quote:
Originally Posted by PRTeamJapan
:confused:
Great! Thanks a lot.
As to mod1.4.... That's why I run it 2-3 times in a row... Usually, 3rd run is the fastest :D
EDIT: BTW, nice clocks and times ;)
Quote:
Originally Posted by PRTeamJapan
Will there ever be multiplier adjustment in the bios for the P4*800 mobo's? That would be so cool because you could then take a 780 Dothan and use a lower multi for higher memspeeds. Has anyone ever tried to adjust a multi in windows?
Nive overclock btw, just make sure to use the msec version in the next run :clap:
the first run is always the best when maxmem is enabled
0.6 better than all the others
fred, to have a regular average, install a ramdisk and put superpi inside
maxmem
ramdisk
1st run
lock pages in memory
LargeSystemCache.reg
Win2k3
priority realtime
optional : iccout.exe and 0.2 better
You are quite welcome & thanks for your tip :)Quote:
Originally Posted by bachus_anonym
Seems you are lucky you get the best time at 3rd run-- somehow I always get worser time when I repeat in a row :(
Thank you for your tips :)Quote:
Originally Posted by cpulloverclock
I know but we will not adopt ramdisk & other tweaks than mentioned in my previous post...
SuperPI is a sort of sacred bench for Team Japan ;)
That Dothan is just insane :slobber:
Somebody should give him some good old Mushkin bh5 or something :D
ramdisk + maxmem = c: + maxmemQuote:
Originally Posted by PRTeamJapan
ramdisk without maxmem is better than c: and no maxmem
Thanks so much for your suggestions :)Quote:
Originally Posted by cpulloverclock
BUT, I still wish to be hanging around at Team Japan as you know a feed back from them on your previous advice to get a better result by kaz-n :D
That is unbelievable. Congratulations!!!
Oh my god, i just HATE when people, who have no knowledge of something new just judge it and fill their posts with insults towards great overclockers.
If people would wake up, go to the Intel section and LEARN, then maybe my temperament wouldn't run crazy everytime i look here :)
EDIT: And to repeat what i wrote when this was posted a few days ago: Xtreme! :clap:
Quote:
Originally Posted by Mathias-K
19s is a noticible irregularity.
Now, I just reread the thread 2x, and have yet to see any 'insults'. People have expressed questions and some disbelief, but I don't see anyone calling him a liar.
surely $900-1000Quote:
Originally Posted by Troman
Quote:
Originally Posted by PRTeamJapan
how do you adjust l2cache in registry????
Thanks Mathias-K :toast:Quote:
Originally Posted by Mathias-K
You have already said all what I wanted to say....
BUT, we, at Team Japan, do accept and respect or sometimes forget about one's opinion though ;)
Please take a look at INTEL 1M at The World Record Database by NiCKE^.Quote:
Originally Posted by urbanfox74
With modPI, already went down to 19s by TAM with his Dothan ;)
Thnaks NiCKE^ and all XS members for your updating the WR :)
Congrats on the awesome score! :toast:
I am considering building a Dothan platform myself. Have you guys at Team Japan tried any of these mobos?
http://www.newegg.com/OldVersion/app...roperty&DEPA=0
Link doesn't work harleybro.
PRTeamJapan, no problem, WRs are also available at www.oc-records.tk :D
It's under the name Dothan 750 but should be Dothan 780 ;)Quote:
Originally Posted by NiCKE^
Thanks harleybro :)Quote:
Originally Posted by harleybro
The link you showed does not work, but AFAIK, only ASUS like P4C800 Series + CT-479.
I am totally blind in Intel World, sorry.. :(
Ohh yeah I copied the html code for the table but forot to edit it haha I forget much things, man it's hard to be old (17) :D
which is the room temp and vcore?
here it's 31-32 and 41C outside
Hmmm hopefully this is better. They are 479 mobo's no adapter needed AOpen and DFI. :)
http://www.newegg.com/Product/Produc...ubCategory=280
No it was actually with Dothan 750 as in hereQuote:
Originally Posted by alexio
I am not sure why this thread was not integrated into the original thread started by cpulloverclock who is constantly pressuring me to ask our Dothan freak member of Team Japan to go over 4G :D :D
I wish you do walk through the said original thread as you can see all their hard work to acheive the results atm... :)
lol...everybody excepting you knows, that this is a known bug of crystal cpu id....
FYI, SecondLevelDataCache in Memory Management at H/L/M System.Quote:
Originally Posted by i found nemo
I do not know exactly how much it may improve timewise as have never tested other than leaving it to 0.
I bet someone knows its secret :D
:D :)
Whats the secret cpulloverclock. I kno u kno.Quote:
Originally Posted by cpulloverclock
sorry about the l2 thing look here i've done the trick with a athlon mobile got a better socore tryed it with a dothan same score.
L2 Cache L2 Cache Tweak
Alias - SecondLevelDataCache Tweak
Myth - "Adjusting the SecondLevelDataCache Registry value to match your CPU's L2 Cache size will improve performance."
Reality - "If you do not set this registry value, Windows will use the HAL (Hardware Abstraction Layer) for retrieval of the L2 cache size automatically on boot up. If this fails, a default value of 256KB is used. This is where the myth that XP can only use 256KB of L2 cache comes from. But it's false. The SecondLevelDataCache is designed as a secondary source of cache size information for computers on which the HAL cannot detect the L2 cache. The HAL is able to retrieve the L2 cache size from any CPU using the set-associative cache design (Pentium II or newer), it's only the relatively few older processors with direct-mapped L2 caches that the HAL will not recognize. This tweak only applies to older computers (Pentium I or older) with over 64 MB of memory. On these older systems, this tweak can improve performance by about 0.4%. Pentium II and later CPUs are not affected." - Source
http://mywebpages.comcast.net/SupportCD/XPMyths.html
Yup, I know you are such a super GURU :) :DQuote:
Originally Posted by cpulloverclock
Quote:
Originally Posted by metro_oc.cl
Thanks metro_oc.cl for your thorough explanation :)
I guess we keep our social customs from the past super OCers....
Thanks again for your valuable pressure :DQuote:
Originally Posted by cpulloverclock
wow 4ghz in a dothan bencheable that is amazing i hope to see some fx57 doing that speed and runing spi 1mb.
Team Japan is Insane!
i really apreciate you,guys
great work and great people
Thanks RoyaL :)Quote:
Originally Posted by RoyaL
Hail to Team Italy :toast:
Quote:
Originally Posted by urbanfox74
"I think that's a rather innapropriate thing to say"
Yes Charlie...With all due respect to you...
That was kinda harsh.
Dothan heats bandwith, if u don't have some high fsb, u don't have a good superpi score, that's only that.Quote:
Originally Posted by urbanfox74
For shure that 3900 with 270fsb is faster than 4000 @ 230, so why u think it's fake?
It's a great honor to us at XS that we have Team Japan keeping us up to date here. Although we don't share a language, I and many others have a tremendous respect and admiration for the Japanese OC "Legends".
C
It is also our great honor to be associated with the World greatest OCers here in XS.Quote:
Originally Posted by charlie
I just cannot say any more but THANK YOU, Charlie.
First time I read about LN2 cooling for CPU was from Japanese website hosted by "Katsuya". He had overclocked Celeron 300A to amazing 716 MHz.
http://homepage2.nifty.com/Katsuya
After that I found amt's overclocking room and his tests with Pentium III Slot 1 CPUs: http://www.horae.dti.ne.jp/~kfuno
I've also followed overclockings of: Ja0hxv, Anpanman, Bunny, Son, Holicho etc. Japanese overclockers are definitely inspiration for my overclocking hobby which I started in 1998 :)
PRTeamJapan,
Can you tell me if the guys are using the new 1022 BIOS with their P4C??
Are they using any kind of shim to protect the core, because mine already has severe fractures from mounting/remounting an evap.
Is there any way one of them could install an x850xtPE and run a 3DMark2001 bench with the card at default settings? We would ALL like to see the results at 3700+mhz.
thx!
C
:D :D :D I certainly believe that it isnt a cheat, and no doubt the japs would even think bout cheating on LN2 to begin with :slobber: :D Nice one charlie :)Quote:
Originally Posted by charlie
Awesome result! Omedetou gozaimasu!! Sadly there are always doubters, even when it comes to some of the best overclockers in the world... everytime someone reachs a new plateau there are those who just can't believe it... Team Japan is hardcore and as PRTeamJapan said, live in their own OC world... but as Sampsa said, they have no reason to fake their results... bushido, think bushido, even in their OC world for them it's all about honor! (^_^)
That Dothan is a BEAST. I really hope it's real, but if according to what charlie said it SHOULD.
Dothans are the future... P4s can say hello to my NBA Garbage tincan...
________
EXTREME VAPORIZER
lolQuote:
Originally Posted by clayton
I hope Intel thinks similar!
:stick:
Hi CharlieQuote:
Originally Posted by charlie
The below is aircooling set up by kaz-n with his 780 ES.
Air Coolinng
CPU M 780 ES
MB P4C800E DX(BIOS1022.003)
Mem KHX3500 256MB×2 400MHz, Turbo, CL2.0-2-2-5-4-E-A-A（1：1 FSB200）
Vcore 1.60V, Vddr 3.45V, Vio 3.7V, Vagp 1.85V
23sec English SuperPI 1M @3049.77(254.15 x 12 with Clockgen)
I am not sure about what kind of shim, if any, they are using..
AFAIK, they are benching at the very edge of breaking point.
As for 3DMark bench, I am afraid that it was politely declined by kaz-n when I asked him before.
I will ask other members but we are not much into 3DMark bench as you may know.
Oh well, there is a old saying in Japan like-- Heaven(or God) will not give you two extreme skills at the same time..
hey I send u a MP for a question about a purchase :)Quote:
Originally Posted by PRTeamJapan
Thank you gocchin for your comment about Bushido-- yes, it is all about what we are based on and are very proud of.Quote:
Originally Posted by gocchin
Are you in Tokyo?
Probably you may know already, when you have a time, please visit Mitsu-san at OVERCLOCK WORKS in Akiba where many our core members are hanging around...
PRTeamJapan/ aka fredyama
what could be said....Quote:
Originally Posted by PRTeamJapan
As requested, 19.141sec by kaz-n.
Cooling: LN2
CPU: M 780 ES
MB: P4C800E DX (BIOS1022.003)
Mem: CORSAIR XMS3500v1.1 512MB×2 400MHz, Turbo, CL2.0-2-2-5-4-A-A-A（1：1 FSB200）
Vcore 1.675V, Vddr 3.50V, Vio 3.7V, Vagp 1.90V
http://premium1.uploadit.org/fredyam...i-by-kaz-n.JPG
Excellent Fred! I see that you managed to convince Kaz-n :)
Thanks guys :thumbsup: Awesome work :slobber:
Oh man that's mighty!! Congrats Kaz-n and Team Japan!! Going to update that..
man ! That's awesome :eek:
Congrats :clap:
Sooo close to 18's! We need that last push to get 18's. U can lower the multiplier, right? If so, tell him to try 16*250 for an ever 4ghz. @ 2-2-2-5, maybe we can break into 18's.
Thanks you all for your compliment :toast:
He said he could bench at 4.05G but the result is slower than as reported.
Unfortunately, he ran out the gas...
Yup, I do hope he will get 18's at his next run :)
Oh man I want that Dothan! Yeah of course I want his knowledge also or I will be stuck at stock speed :D
R e s p e c t ........"it's all about honor"..... :) :clap: :clap:Quote:
Originally Posted by gocchin
I think he's just "show up" a bit for leting the others imagine what this chip can do in a few days.......Something like a "warning" I suppose...... ;) :DQuote:
Originally Posted by charlie
Wow congrats on the score :toast: looking forward to seein 18.
Wow, very impressive.
wow nice
FRED, can he test a max screen?
4.2GHz for example :D
Congrats guys, amazing stuff :)
I should apologize for saying it might be fake, my original statements of disbelief, where because I was miss informed. I am not to failmiler with japanese ocers, So unfourtunately not many names of there oc/ers I reconised. :slap: Ah well, glad its all sorted out and really happy that its real thats some awsome power there. Much respect indeed. :D
Very Impressive :D
Let's see what my 770 will do :D
Nice clock, nice score. Almost getting to the point of that supercomputer the original authors mentioned in their readme I think.
What happened to Holicho ? He was one of the few Japanese overclockers who liked to do 3dmark.
Regards
Andy
Thats incredible :|
congrats :clap:
Insane clocks on that dothan :toast: