Clevo Overclocker's Lounge

I didn't do much to it except of adding bunch of heat pipes to keep it cool and running a custom vBIOS.

 

Try at your own risk lol.
Dont want people coming after my ass. It works best for desktop cards
Also it stresses the VRM's a lot so i wouldnt suggested it unless the GPU for sure is properly padded and cooled. Heck i wouldn't even do it for general use, unless you know exactly what you are doing.

 

Well, actually got around to finally testing SLI. Things seem promising with a bit of headroom left...Some fun runs.
And no shuntmod!
http://www.3dmark.com/fs/12776057

 

This makes one think and wonder what does the future have in store for us.

https://www.facebook.com/video.php?v=1324595060970269

 

Dark Souls: Prepare to Die Edition, lol.
I am ready to burn my card since the first day I got it

 

Maybe. You might be able to hex edit the SPD firmware as well. All I know is I don't find it in Thaiphoon Burner.

 

A little better run here.

 

Next will be +P34000

 

P34000 And 63K gpu

 

You beast!

 

dunno how reliable Testmem 5 is, cause " stressapptest" rekt my NB pretty hard so i used the first one instead but i think after you guys use it i guess it's fine



too bad tREF resets to 12480 everytime , if i set e.g. only 12500 it's in bios, reboot boom 12480 again.

 

I think I found the issue LOL, turned on windows classic theme....

http://www.3dmark.com/3dm11/12205861

 

Nvm, I guess that wasn't the issue..Tried running three more runs on each theme to see if there is consistency but the score went down again.

http://www.3dmark.com/compare/3dm11.../3dm11/12205919/3dm11/12205924/3dm11/12205861

 

TM5 is pretty reliable, comes very close to the golden standard (overnight memtest86+). TM5 will get you to about 99% stability of an overnight memtest run, definitely sufficient for everyday operation

Sent from my HUAWEI NXT-AL10 using Tapatalk

 

+150/300 on GPU & 5.0Ghz CPU

30k 3DM11 & 27k FS:

http://www.3dmark.com/3dm11/12205934

http://www.3dmark.com/3dm/20263675?

 

hmm maybe no stable 3200 for me, i did a 30 min stressapptest successfully, then i turned off my NB, did the test again 50 errors in some seconds, then i closed Testmem5 let it stay for 1 min opened again and started stresstesting, what happened was that the app was running but the background processes silent crashed :/

 

http://www.3dmark.com/fs/12786073



http://www.3dmark.com/spy/1853870

 

ha, approaching 3x FS score compared to my 980M xD

 

haha yeah, i don't look much better either with a 1070, these guys are OC heroes, and @jaybee83 i know now hat you mean with the ram oc thing is a "depends on her mood", some boots it's stable and next restart everything is going down again even tried some Uncore +20 +30 +40 +50 and so on even -20, maybe it's that 6700k thing + mainboard + whatever combination i read when digging through some threads, right now tho i'm testing 3100 15-17-17-39 which right now is stress testing and looks good so far . I know Intel scales much better with speed than latency, but still going from 3000 16-18-18-43 stock to 3100 15-17-17-39 seems like a bit weaker 3200 anyway in terms of bandwith and latency.

 

http://www.3dmark.com/fs/12787801

 

Looks like your 3DMark has auto-updated itself and cost a bunch of GPU score...

2777Mhz...as a company that does almost nothing but benchmarks, they should have figured out a way to read actual boost clocks from sensor data rather than relying on guesstimates...LOL

Edit: BTW, you guys are OC MONSTER!

 

Yeah, it updated itself without my permission. I have been hitting the ignore button ever since the latest messed up version was released. It is still telling me updates are available even though it already updated. I guess I need to uninstall, use Revo and CCleaner and start over again.

 

Interesting.... I'm using the same version as well
http://www.3dmark.com/fs/12788955

 

Maybe it doesn't have the same impact on old driver as it does on the current ones?
Or it is just a part of the update? Many user reported a bug which caused a rather large drop in FS GPU score with all newer 3DMark versions than 3509...

 

Not sure. I was thinking it was because he was running 3200 mhz ram while the rest of us were at 3000.... Speculation because my score only went up by 110 points or so on combined

Also he has a much higher combined score which seems to be far better than having a higher gpu score for this particular bench.

This one is just up for grabs i guess...

 

Yeah, and I am using a new driver right now. Maybe they fixed all the problems it had quietly, and without a version change. I had trouble with it before having not only lower scores, but also crashing to the desktop and saying I cancelled it when I didn't.

 

You are totally right about the versions being jacked up. And to be perfectly honest... I was thinking this version was jacked up also. Have no clue why it actually worked. The first two runs were 47.5k then the 3rd and 4th run was 48.4k. So i'm going to chalk it up to being an iffy version at best, but when you crash too many times with 509 it corrupts in a way that is very hard to recover without doing an os reinstall (And hard to reinstall any other version as well). Or at least for me that is...So trying to have a backup version that will run is not a bad thing.
And in some cases 2067 is better.

@Mr. Fox Your brightness still works right?

 

I have been trying to find 2067, will you be able to upload it somewhere?

 

*logs on, sees score, thinks **these dudes are insane** likes everyones posts, continues watching*

that could be the culprit i had maybe when between non-shunt mod and shunt mod, the first benchmark wasn't too impressive but lost about 700points both @ stock, so not a Windows 10 CU issue i thought. Thanks for the headsup

oh and 3100 15-17-17-39 runs stable @ 1.25v did 1h stressapptest with 0 single errors, rebooted several times + big shutdown went to lunch came back stress tested again, still fine, so yeah i guess that's some voodoo thing why 3200 isn't running (always).

 

It will take a minute to sync with google drive....

 

You find all older 3M suites for download on Techpowerup.com

 

Hmm i didnt know we could find them there. Most other download sources are either too slow or corrupted. Thank you though! FOr now @Johnksss@iBUYPOWER to the rescue!

 

That's where I get it, too. You can get a lot of older and better (seems like sad a trend) software from TechPowerUp.com web site.

 

You should backup your 3DMark suite as main rule. New crippled version will come as standard now. Especially around a new Windoze X main build Micro$hit have made a new trend. 2 times a year and Futuremark can't follow up

 

I have all of the Futuremark installer files that I use stored locally, and yes... very good idea to do that. I fully expect to see manipulation nonsense coming from them as well, like capping scores on certain hardware IDs like the super-strong Clevo notebooks to keep them from raping BGA turdbooks and leaving them to bleed out in a ditch. I would not be surprised if we are not already seeing this happen to some extent. I suspect they will gladly accept bribes from Intel and NVIDIA to do dishonest things to manipulate the numbers and make the BGA feces more palatable and sellable. We live in a time where being crooked is tolerated where someone stands to get rich, and I have no qualms saying that I think all of the big names in PC hardware tech are worthless, dishonest, lying bastards that don't give a rat's ass about customers or doing the right thing. None of them have an ounce of integrity.

If they were decent and honest companies run by good people there would be no such thing as these syphillitic BGA "gaming" fecesbooks because they would not allow their brand names to be associated with such utter rubbish.

 

That's quite a serious accusation to say they are manipulating the benchmark to favour one company over another.

 

I would not find it surprising at all, but I agree that it is serious. Not any more grievous than the other manipulation and nonsense that goes on with cancer firmware and driver voodoo though. Equally unethical in my view. I don't think we can trust anyone that takes our money to do the right thing at this point, particularly where mobile computing is concerned.

 

In the past they did not bend to them, it would be sad if that stopped being true.

 

That looks pretty sweet. Are you the author of that tool?

Yeah, I agree with you 100%. It would be really sad. But, having a basis for sadness seems to be the norm for notebook enthusiasts now, with but a couple of exceptional (perhaps even accidental) examples of goodness.

 

Yup, . The current version can't remove the Pwr throttling completely. It throttles the 1070 around 200W and the 1080 around 240~250W.

 

Yeah, it's definitely going to take @Prema Magic to exorcise those demonic elements from the firmware. He is very hard at work on doing that, but the powers of darkness are strong in Pascal.

Anyhow, awesome job on that tool. I am amazed by your talent. It looks like it would work with the stock vBIOS fine.

For those that will never enjoy any of the benefits of a Prema Mod, this would be better than the mediocre existence that is representative of the norm today... a real godsend for the gigantic group of unfortunate folks that represent the norm.

Edit: @Coolane - http://www.overclock.net/t/1621789/pascal-bios-editor-any-news/50

 

Maybe one day an updated version can remove the power throttling completely.
I was testing some addresses in the 1070 trying to find the evil address, lol.

Thank you, man. Glad to see the tool was also introduced over in the overclock.net.
I was thinking desktop cards can cross flash vbios from different manufactures, that's one of the reasons I didn't make it work for desktop vbios. Also, I don't have a desktop card to do testing now.

A lot of 1070 users still have their TDP locked down to 115W, hopefully they can find the tool useful and release their power.

The burning MXM logo looks cool, right? It took an afternoon drawing it with 10bit color,

 

That niko2004x really doesn't know what he's talking about. Throwing guesses around.

 

I was hoping you would see his posts and share what your experience has been with vBIOSes for desktop GPU(s).

Yeah, looks nice. Good job.

 

I tend to not reply to most people over there. They are bit too thick skulled to listen logic and testing.

 

Yes, but we could say similar things about the BGA Kool-Aid drinkers here now. I used to never engage there because there were so many laptop hating trolls. Now that almost everything is BGA filth, I think I'm destined to become a laptop hating troll myself. The only difference is that I will have a legitimate basis for the hate, and they don't. The P870 has changed the perception of mobile trash with many of those trolls. There are not as many haters as there used to be.

 

Hi, thank you, but we can't flash the modded Bios without a IC programmer and removing the Bios chip, right? NVflash is impossible?
(im using a GT73VR with a GTX 1070+7820HK. Access to a 330W PSU).

Why are some screenshots with a red checksum and some with a green checksum? What is the difference?

 

That thread is going to have some very interesting posts soon. XD

 

You do not have to remove the BIOS chip. You can flash in-line with a 1.8V adapter and alligator clip with the TL866A programmer.

I'm not an expert on the checksum thing, so I will let someone else answer why it matters. All I could do it give you a link to wikipedia rather than respond with something meaningful in the context of this discussion. Maybe @Prema or @bloodhawk or @Coolane can speak to that in a more professional manner than I can.

 

Some programmers can inline flash it . A lot of caution is advised.
When it's red, that means checksum is off and needs to be matched.