Very strong and efficient run!

Well said bro Keep up the good work Johnny!!

My God!!!6.9 and 10.26?!!!impossible! what happen if you run @ 6.95! congratulation johny;)

Monster CPU!!! - Nice John Lam...

Oh Personal Best

Edited October 3, 201212 yr by Leeghoofd

Impressive RAM! And CPU ofcourse...

welldone

Spill the I-ram trick !!! Or was that for the worlds fastest PCmark run by CherV ?

You benched together John ? teaching her the tricks

Edited October 4, 201212 yr by Leeghoofd

PI-Fast tricks and tweaks were there...

 

http://www.hwbot.org/forum/showthread.php?p=201236#post201236

 

Check the screenshot , some error happened in my hide mode... I will rebench it later.

You have uploaded at least two different screenshots to your 10.26s submission before selecting that it should be awarded no points.

Here they are both still in hwbot database:

http://img.hwbot.org/u41156/image_id_847393.png

http://img.hwbot.org/u41156/image_id_847721.png

On the first one, there is no space between "0.89" and ":" in the division time.

Second one is even more badly messed, having "8.89" instead of "0.89", ~0.01 hours next to 10.26s and a letter "g" cut off by the "ASRock Timing configurator" icon.

 

If this isn't photoshopping scores, then I don't know what is.

Edited October 4, 201212 yr by TaPaKaH

We are aware about the matter, it is under investigation.

@ Sam OCX

 

Be Honestly , I feel sorry to everyone because I haven't double check my result.

 

I have upload 2 screenshot because massman told me my screenshots have some problem and I upload the second one without checking.

 

I can say this is unbelieveable, but MS Ramdrive is very buggy and the system become very unstable.

 

This is the first time I tried this method and support will not have any problem with MS Native RamDrive , please forgive my careless.

 

Sorry X 1000000 to everyone and this is not cheat

even the wprime 32m, I really afraid this is a fault run , so I want to make sure my tweaks and trim is work.

 

So , Yesterday I make a test in 3960X

 

before tweaks (3s)

 

after tweaks (2.921s)

 

this is my careless , even the wprime 32m result , you can see my careless style. Sorry T.T"

if you still have the exact OS you did PiFast(s) on - can you show any SuperPi calculation at any frequency with any level of optimisation?

this is not a result Cheated - John Lam work and study hard to get it, it was just an oversight on the screens of CPUZ

Now tell me who never committed any negligence

can not judge a person just because she committed an oversight

There are many errors in the league world overclocking

I mean that if Andre Yang - Shammy make a mistake like this they will be accused of cheating??

Is not that how we have to look at this situation

let's have a bit of reasoning here

he is only seeking new tweaks

we will demonstrate respect for the work of this man

It's under evaluation by the staff, await there call.

 

FYI there's nothing wrong with the CPUZ's by the way... other things look suspicious...

john lam if you get banned for a year can I have your chip please

If you cheat at imaginary computer points you cant argue that your a douche.

john lam if you get banned for a year can I have your chip please

 

.... What Price ?

.... What Price ?

 

marital rights to my sister?

If OS is indeed the same, then it is normally not possible to have xx.26 seconds as a PiFast time or subtime.

 

Explanation:

Normally, PiFast counts time in 1/64s of a second, same as SuperPi.

16/64s = .250s in SuperPi = .25s in PiFast

17/64s = .265/.266s in SuperPi = .27s in PiFast

Try looking for 11.26, 12.26, 13.26, ... results in PiFast on hwbot and see how many of those were done using XP.

 

The only way to have .26s on XP is by having OS messed up in a very specific way that both Pi and PiFast will not follow the "1/64s" rule. I remember Hicookie had something similar in Clarkdale times when his records were questioned (5.910s 1M on XP or something like this). From my own experience of nLiting hundreds of XP versions, this bug can emerge only during OS compilation and is permament. Now that you've shown via SuperPi that your OS doesn't have the bug, I have more ground to think that your 10.26s result(s) were bogus.

 

To staff - I'd really hate to derail any tweaking thread, can the "offtopic" posts be moved somewhere else? For example, into the submission discussion thread.

You told us that TWCL setting is very important for PiFast and you have told the same about PCMark 05. The last think I want to hear is that is important for Wprime too. After many years of benching I believe that we now how to manage mem settings.

draw errors could affect result too and Sam OCX explanation made me believe that John's pifast submission can't be accepted.

Same steps are for Wprime too different for each OS and i'm not sure about 3.375 for WinXP. Results with X79 are normal but with 3770k?