Printable View
http://valid.x86-secret.com/show_oc.php?id=309749
Attachment 72601
It's fully valid according to Franck but no email addy etc. :)
So, who are you kris, how did you achieve it, specs, details, pics, any benches, coldbug, etc?
We're talking C2D MHz here.
Official hunt declared. :cool:
I already post in the x2 overclocking and discussion, but no1 goes there:(
I have doubts about this clock...thats a stunning 587,4 MHz clock difference to the second place. And its done on a AN-M2HD with IGP If thats real im very impressed.
Hmmm... I'd say bug :shrug:
yea we need a setup and volts ASAP. Nice work. Looks like I need to order another 5000 BE, mine is walled at 3300 :(
uh, K8 just can't clock that high :x
Obviously it can. Only question is how did he do it?Quote:
Originally Posted by biohead
Probably a suicide run and likely no benches at that speed. Great overclock though! :D
I was thinking the same thing, 1 run wonder!
^^ his name is chris too. omg1?!?!?!
Yea I figured that too.Quote:
Originally Posted by freeloader
The highest I can get right now is 34xxmhz @ 1.45v load into Windows but then it BSODs. It sucks. I'm thinking about ordering another BE and see if I can get a bit further. I really think I can hit 4Ghz on water and make it stable.
vote bug/generally BS.
Arent there only a handful of people to even get AM2 over 4GHz, now all of a sudden we have 4.8GHz? Yea, right.
Some 2.5V vCore and some crippled Windows boot, fast validation and crash. That, or fake. Go figure.
2.5v on the cpu would be nuts. I'd expect mine to start smoking or something.
Well, bug vs. no bug is a tricky one. There's no CPU-Z bug according to the man who verified it and verifies all of our CVF files; Franck deems there's no difference in this validation and any other X2 genuine one. But I'm still thinking how and also, the guy who did it must have known what he did was a one off...
In which case he would've boasted about it somewhere at least.. if it was for real, since the cooling and setup required had to be LN2 aswell :wth:
Not sure what happened here, time should tell.
It's either :bs: or :wierd: or just :banana3:
wow...on a crappy mobo to boot...and both cores are enabled either someone got lucky and found the wonderchip or cpuz has a flaw!
CPUz shows wrong frequency very rarely, but it happens. This is (obviously) not a real score. First of all, it's just too good compared to other scores (like a 7GHz Yorkfield). 2nd reason: AMD chips have a different coldbug than Intel chips. Lower temps = lower FSB limit (not a wall, like in Intel chips). 300mhz++ fsb on AM2 can't be achieved on dice or ln2 because of this, and it's not possible to get that score on something like a chiller, no way.
And can that mobo even do that high FSB??
4850mhz on AMD it's actually hard to belive:rolleyes:
AN-M2HD isn't that crappy especially if you don't use the IGP & improve the chipset cooling.Quote:
Originally Posted by cdawall
It's certainly capable of doing 323fsb.
I think I read somewhere on these forums that it's possible to fake 3dmark scores, and high frequencies.
found the thread:
http://www.xtremesystems.org/forums/...ad.php?t=46926
If it's real, even just a suicide run, it's still impressive.
I have seen this in the X2 thread btw
Maybe he got past the cold bug? Dont know if it's possible, but if you make sure the CPU stays around lets say -10 and just increase the cooling every Volt/Mhz bump so it stays at -10?
So not a -200 run right away, start slowly and adjust the cooling to target temp?
With voltage needed to achieve 4.8Ghz this boards pwm would blown up. G2 Brisby have awfull coldbugs, so this is totally faked result, there is no way to reach 4.8Ghz on this chip. Even 3.8Ghz on this board would be hard to believe.
wow............it's the highest oc to amd cpu i saw !
kewl and great news ! :D
Not so great when it is fake. :shrug:Quote:
Originally Posted by gillll
Seems this Kris is from Nordic Hardware and had some posts there with this chip before, making it into top 10 lists on that forum easy. It might be a fluke, but I wouldn't mind if it weren't :)
If I see superpi1m at 4.5-4.6ghz I'll believe it.
Linky: http://www.nordichardware.com/news,7362.html
although no real info :|
fake.
No way. That's really impossible.
Fake.
Fake
/ bug
Well, well and.. well. :D
Have a look at this from the 4th of Feb: http://valid.x86-secret.com/show_oc.php?id=308020
http://img90.imageshack.us/img90/3204/42ghzyz5.png
Check his previous results out: http://ripping.org/database.php?username=kris
http://img205.imageshack.us/img205/9733/kristianea9.png
I'll keep listening to the comments for now. :D
Regardless of that entry, not all G2 Brisby's are CB'd, mine wasn't at least to -30C and did 1.3V 3.52G on stock HSF air. Temps were the limit really, idling plus 50C.
What FSB did you run at when going -30c?Quote:
Originally Posted by KTE
Let's say that his CPU is coldbugged at -75C ( K8 with cold bug at -75C... keep on dreaming :D )... that temp won't be "good enough" for 4.7GHz + clocks...
There are several bugs ( for example remember the nF680i bug with the first two versions of CPUz back then ? ), some of them were discovered and corrected, but not all...
I'd like to believe it, but it's nearly 100% impossible.
A bench ( 3D Mark2001SE for example ) at 4.5GHz would be impressive as well... but I hardly doubt there will be any ( and this CPUz validation is more than likely fake )
Too : http://valid.x86-secret.com/show_oc.php?id=308132
French flag on RIPPING.ORG but he is unknown* on french OC forums!
* PCimpact user, 2 posts...
world record :up: congrats!
Don't have the HD with images with me here but from memory, IHS temp -30C was my stock Vcore bootup (208x16).
He is french so XS french guys might know him and ask for details
Congrats with wc, wouldn't call his results fake until proven so..
So, we have a insanely high score of a person no one even heard of and no information or pictures anywhere?
Looks to be an easy call ... fake :).
This is the guy who knows him:
http://infomars.fr/forum/index.php?showuser=17
http://forums.vr-zone.com/member.php?u=59052
Ask him. ;)
He is know as kardiakkris :
http://www.pcinpact.com/forum/index....pic=120343&hl=
I read the thread now !
EDIT : OK, hum... I don't know how he do that but it's not a real frequency... hard to explain in english...
#62 : http://www.pcinpact.com/forum/index....c=120343&st=60
VIDEO 1 : http://www.megaupload.com/fr/?d=UTZ7TH1J
VIDEO 2 : http://www.megaupload.com/fr/?d=R2UK0BJ5
He admitted it's a fake.
It's fake, but apparently not because of a hack nor because of a bug. As far as I can understand the explanation, he says it has something to do with Cool&Quiet and NB/SB cooling.
TBH, I don't think it's worth the effort of finding out how he did it, really :).
Agreed.Quote:
Originally Posted by LIKMARK
Someone with an intel chip reaches an insanely high speed, everyone goes "wooowww....."
Someone with an AMD chip reaches an insanely high speed, everyone goes "fake!"
:rolleyes:
whatever.
Until proven otherwise I'd be led to believe it since it isn't just a screenshot but an actual validation.
Maybe you guys missed this: http://valid.x86-secret.com/show_oc.php?id=309280
http://img232.imageshack.us/img232/6217/hahari0.png
Over there they're accusing him of being a cheat, a fraud and hacking CPU-Z but he said he did not do anything as such and just did some air cooling oc. He said everything gave him the same result, even using Clockgen he could get it.
What is he saying? He's saying there are slight FAST fluctuations in core speeds and you can capture those as valids, even in EVEREST. This happens even on Phenoms. Check his videos out.
He looks to have exploited the weakness in oc's I know about for a while, and so does Franck and most software devs and unfortunately there is no real way against it. If a result is far off, then you can spot it sometimes but if it's near what others are getting, you'd call it genuine while it's not. You'll believe you have a fast result while it's just a bugged reading, it's basic electrics I warned about long ago. And you can't have it unless you use SetFSB/Clockgen sort-of applications, since they screw with the core PLL that much.
He could probably very easily fake higher MHz and get higher benchmarks.. as long as they're synthetic.
Everyone would be "fake!" if someone got 7GHz with C2D. If this was some 4.2-4.4GHz then people would have been like "Woooow!".Quote:
Originally Posted by SparkyJJO
And btw. it was HACKED. He hacked it! HACKER. A REAL one, not one of those kids with some trainer software. Respect those. :)
Grow up...Quote:
Originally Posted by Calmatory
Excuse me, but if he actually GETS better FPS/faster time in SuperPI due to incresed CPU speed, is it a cheat? The idea is to get the system to run FAST, thats what he is doing. Sure the CPU-Z and alike are exploitable, but if that thing improves SuperPI due to "cyclic overclocking" then sure it is a good tweak/hack. I'd use it and I see no problem with it, if it REALLY speeds up the CPU.Quote:
Originally Posted by KTE
The idea is to et FAST machine, not stable. Ban suicide runs as you can't use them 247 and they have nothing to do with real-world performance!1!"
Edit: And to add, that IF the core PHYSICALLY ran @ 4.8 GHz, the result should be valid, obviously. But if the software was fooled, then not.
...if it is just something which fools the software then let that idiot burn in hell. :|
Why? Hacks are nice, really, as long as they aren't used for exploiting stuff. In this case it was, but seriously, there is NO ONE else who could do this? Can you say he should get no credit for this? Let me repeat: No one else could do this.Quote:
Originally Posted by cozwin
And remember that there IS a difference between hacker and cracker. Too bad that e.g. media has fooled people that nowadays they hate hackers, for the stuff crackers do. Like making people to hate police for stuff what criminals do.
QFT, my nView (the older nF4 brother) does 325 24/7 with no extra volt on nb with my Brisbane.Quote:
Originally Posted by Heidfirst
Did you read what I posted above? These fluctuations are common on all CPU's if you screw with the PLL's, they are not fixed speeds nor the core speed, they are slight millisecond fluctuations. Some cores fluctuate more than others. If you can have any application which monitors and picks these up, you will get them registered. So benching them with anything real will not show any benefit whatsoever, only synthetic software which won't run real tests will show higher results if it picks up higher MHz and those MHz stay consistent. These MHz don't. Synthetic software such as EVEREST/Sandra/CPU-Z Latency gives you results based on the speeds it detects, not on the real component speeds.Quote:
Originally Posted by Calmatory
Super Pi isn't synthetic and it won't show any faster time, nor will WinRAR, nor will any rendering such as Cinebench. ;)
If it did, then the result would be valid, as the core must run quicker than 4200MHz to beat a 4200MHz time by a margin, quite obviously. There is no hack involved, none needed.
So I was mistaken, and all this stuff is well knon from before, now soeome just decided to exploit it, right?Quote:
Originally Posted by KTE
Hope he gets banned from everywhere. :|
Check your PM, I'll explain with evidence. ;)Quote:
Originally Posted by Calmatory
I saw the CPUz freq bug myself during some testing. Lasted only for a split second, so I was unable to save a validation file... But I'm sure it was like 100ghz (could be 1thz as well:D )
This bug is easily spotted, because the FSB is the one that goes up (NOT the multi). And thanks to mr. coldbug it's impossible to get those scores in a "normal" way.
Quote:
Originally Posted by Marvin_The_Martian
He's not from NordicHardware, just reported on the "achievement", which is apparently just a PLL exploit.
//Andreas
Sorry for the confusion seemed that way at the time.Quote:
Originally Posted by NH|Delph1
No harm done, it actually reminded me to update the news :)Quote:
Originally Posted by Marvin_The_Martian
//Andreas
I Know where Kris is :p
:D :D
dont be so desperate for a new AMD record, that you will accept a record with no proof, name, face, results, or anything else.
just cause its on CPUZ doesnt mean its valid.
it means he found a new way to exploit CPUZ, nothing more.
for all the effort put into CPUZ, you cannot make something cheat proof.
only harder to cheat on.
franck does great finding ways to keep CPUZ as reliable as possible, but theres always a way to fool the program, somewhere.
this is farrrrrrrrrrrrr to fishy to buy into.
so anyone here able to translate that french?Quote:
Originally Posted by Kyro38
link from OP shows that its now rejected
http://valid.x86-secret.com/show_oc.php?id=309749
if you click http://valid.x86-secret.com/show_oc.php?id=309749
and read the first line it says REJECTED
Already posted, page 2 :DQuote:
Originally Posted by Kyro38
[French] Si il pouvait donner des informations en anglais ICI ça en interesserait plus d'un [/French]
next time, read the last post. i said the same thing roughly 1 hr 40 minutes earlierQuote:
Originally Posted by godsa
lol some are even congratulating him in this thread :shocked: LOL!
well at least cris make me shock and almost buy the same setup to try :ROTF:Quote:
Originally Posted by dinos22
lol this was such a giveaway it's not even funnyQuote:
Originally Posted by R*r
how could anyone with half a brain believe the CPUs are capable of such clocks when we all know they struggle to break 4Ghz in most instances lol
and then 5.5GHz hahah
you guys would have seen those crazy bios CPU speeds that Asus A8N SLI produced with some bioses showing 6GHz+ hahaha
Aww cmon dino you guys are having trouble cracking 4 gig?
I may have to throw my 939 rig back together see what my cherry x2 3800 can really do, remember I just cracked the record of 3505, with 3508 for fun......looks like 3700 is the # to beat, If manta plays nice with ram and i use a better DI tube :D
chew
amd is a bit of a joke at the moment
i wouldn't buy any of them until they release something half decent
It would seem, that ANY scoring done with CPUZ would now be suspect.
If all of this IS fake, then apparently the CPUZ results are suspect and ALL the scores and claims of super speeds should be suspect. This score of Chris' was validated at first.
and BTW, what is to say that this bug discovered, may have been present for some time, and shared and exploited by the "intel is better than AMD " crowd.
Kind of like the BIAS we heard about with the NVidia vs ATI GPUs under, what was it, 3DMark ?
Sure sounds that way to me.
You mean like the Phenom at 6GHz on air stock volts? :lol2:Quote:
Originally Posted by dinos22
Had my X2 6000+ at over 7500MHz my Crosshair bios told me lots of times... :)
Yes, i've seen that a few times.Quote:
Originally Posted by dinos22
A while back i also had a similar problem on a M2a-vm board. The cpu ran underclocked but showed up as 3.8ghz. I also had a superpi 1m score of 7.6seconds lol :rofl: :ROTF:
Lol what does cpuz and the "intel is better than amd crowd" have to do with anything........Intel is not breaking amd records, they are not reporting 5 gig brisbanes with conroes....via a bug......Quote:
Originally Posted by Strider7
Also atm, intel just clocks higher and its common knowledge, there is no need to exploit......I pull off a 800mhz 24/7 stable Oclock with stock volts on a quad......I personally don't like intel as a company.........However I will not pay for for an amd quad till they get back to they're roots........etc beating intel performance in the clock for clock department and having a high yield.....
very big FAKE, my athlon 64 2600+ is NOT fake. :)
Still waiting... :DQuote:
Originally Posted by KTE
If you check out the thread he started in the bug reporting section of the ripping.org forums you'll see that he claimed stock cooling/1.35v :clap: I can't believe some people actually believed him, truely I can't...
Let him pass validating with the latest version of CPU-Z 1.44.1 Still no evidences.
4.2 ghz is a bit closely to truth.
it is, but if you read earlier in the thread, kris has apparently said that it's a fake tooQuote:
Originally Posted by Data_
looks like my chip is still the only 65nm amd to have hit 4ghz? i find that hard to believe :confused:
http://valid.x86-secret.com/show_oc.php?id=266059
It's a system bug rather than fake. ;)Quote:
Originally Posted by ozzimark
There's many that should hit close but most don't try more than air. Pretty sure mine could've gone close, did 352xMHz at 1.4V stock HSF air. Needed more cooling, didn't have it then.Quote:
looks like my chip is still the only 65nm amd to have hit 4ghz? i find that hard to believe :confused:
http://valid.x86-secret.com/show_oc.php?id=266059
Heh, I will put my 9850e up against your E2140 any day in any application. :DQuote:
Originally Posted by dinos22
Haha, I love it! People like this make my days more fun. Seeing title, not reading what have been said at least in the previous 2 posts and posts the comment of the day. I love it, lol.Quote:
Originally Posted by gillll
Missing you KTE... :(
heheh i missed your little "dig" hereQuote:
Originally Posted by bingo13
you know what i mean
extreme cooling etc
anyways Ph2 is coming out and in the benchroom about to get frozen on sunday so clearly i've moved on and found something useful in AMD to play with hehe :D
I don't really understand how my method works. Must say.
I'd say do it, then CPU-Z will read this strange core speed.
.cvf files are not always validated but some go thru. No matter the 1.4... aso 1.48 versions.
Question is : CPU-Z calculation wrong ( refresh rate maybe ) or real instant core speed ? Some programs are able to change FSB instantly from OS to system. If the rate is not fast enough which process is prior ? Time to gather the various pole points or time for the PLL to compensate ?
I did not intend to annoy the fame of the sub-zero ways to reach high Core Speeds.
This is only overclocking. This is not science.:)