Clevo Overclocker's Lounge

LOL

EDIT- @Papusan Do not fret good sir. Good things come to good people.

 

Meanwhile, back at the ranch.. getting closer.

 

Man i cant wait to get me second 1080 back. Hopefully it comes in tomorrow. Otherwise this will be a boring weekend

 

Well, wasn't a boring Eve for brother Fox, that's for sure, but I wouldn't use the word fun either...though I think he's ready now for the job at the Clevo factory assembly line.

 

I have to wait till Tuesday on my keyboard. Also, going to have my nephew who's here learn some things...

Sent from my SM-G900P using Tapatalk

 

It looks impressive,
so far this is what I understand from the video:
I'm positive the computer on the right is a DM3, there's a white box in the middle that's attached to one of the black squares on the video card and that's attached to another computer on the left, (make and model I'm not sure) Mr. Fox is definitely doing something with the Vbios, (I got that info from the title of the video) other than that I'm lost. If any members are as technically savvy as I am? Don't try to flash your own Bios, if at all possible buy, from a Prema partner. I should copy and paste this to the Tornado F5 forum

 

100 % sure the another computer ain't a Jokebook. Or what he would call Filth!! But the EUROCOM Tornado F5 Mini-MonsterBook

 

Just noticed that!

 

so, propably a bit stupid question but it urges me somehow, does the amount of Conductonaut matter how much the shuntmod get's me in power? e.g. i just played WoW on my book and hooked my new screen at it and played some hours, and noticed i still get in the power limit sometimes, ~1925mhz vs ~1650mhz back then using 1v and 1950mhz setup. Don't get me wrong it's hell of a difference in clock stability and OW seems to profit quite big from it and the shuntmod adds about 20-30w (?) extra, but back to my main question, does the amount matter?

@bloodhawk

 

From what i have heard and observed on my own, it just might. I did not personally try adding more, since the vBIOS mod was working way better for me.

 

ok thanks, till i can do some vbios mods myself i just leave it at 987mv @ 1980mhz, i will fiddle around later with it then but so far 122fps OW capped is quite a change prior to the previous 90-120fps adventures. Too bad the newest nV driver is **** and reintroduced the gsync blackscreen flickering

 

Yeah thats the safe way around. Im really not comfortable with too much liquid metal on my GPU's lol.
As for the drivers im still using 381.65 on my desktop and 378.66 on the laptop.

 

yeah will roll back too, Gsync was actually quite a nice bonus after i learned how to tweak it to perfection, simple said OC the WHQD screen to 125hz and fix the fps to 122 feels quite nice, too bad gsync in notebooks are borderline working, on my desktop after i got my gsync screen 2 weeks ago, everything worked out of the box, still it isn't a "must have" for me but for ****ty optimized games like GW2 it works wonders.

 

here something to post, new highscore

CPU @ 4.5Ghz GPU @ 2000/4500

 

Definitely not boring, that's for sure. I lost count, but I think we tested something like 18 or 20 mods last night. My Alienkiller needs to be held together with Velcro, zippers and snaps instead of screws, LOL.

This is with the GPU stock. Check the wattage. It now seems like it runs stock the way it should have run stock if Clevo had actually done things the right way. Will have to fire up the AC and see what this bad boy can do.

 

See? I said it didn't I? 220W on the GPU is enough to hold full stock performance on these things. 190W ain't bad but 220W would have been perfect for a standard release.

Of course you and the other benchers here aren't "standard", and mods for you are great, but I think this should've been the initial release design.

 

I agree. I'm sure the reasons it was not are about as silly as the day is long, too.

 

My IHS should be shipping too

 

720W on this run...

 

And, bear in mind that is what HWiNFO64 captured with 2 second polling. It did not capture the spikes to 765W that I saw on the screen of the UPS meter. But, it's not in the screen shot so it didn't happen, LOL.

Edit: the memory reporting in HWiNFO64 is also wrong with the current mod. It's actually 1500 now (my previous max was 1400 as shown in GPU-Z).

http://www.3dmark.com/3dm11/12198073


@Johnksss@iBUYPOWER @Khenglish

 

Stock run, scores seem low:

 

Nice!
1500Mhz (12Ghz??!!!) on the memory is pretty insane for a 1080 IMO!

 

@dspboys http://www.3dmark.com/fs/12740341

 

Yeah, he didn't intend to. It was done in error due to me forgetting to adjust the gpu-z report value. He thought it's running a couple hundred Mhz slower on that first test run there...

 

That's from the wall though, right? From the 780W brick it should have been a bit under 600W?

 

The Adapter measures on the DC load line. AC pull from the wall will be slightly higher due to conversion loss.

 

Don't expect more than 90% of the power pulled from the wall

 

I installed mine yesterday. The Bitspower Skylake model and my temps are a little higher than before. I don't know if I used too little CLU, or some other bone headed move. I'll have to try again.

 

If you got rid of your AC-100 adapter I have a spare . Looks like you guys will be using #2- 780 watt soon. WTH!

 

i had the same at first, did another reapplication of clu and then i had better temps with the bitspower ihs. check and see if load temps of your cores are close to each other, if they are then thats a sign for a good application mine differ max. 4C from each other.

Sent from my HUAWEI NXT-AL10 using Tapatalk

 

Thank you, I appreciate the info. Mine are all over the place, they differ more that 4c. I was worried about getting CLU on the small part next to the DIE so I used very little. I'm not positive I installed the IHS correctly either. I didn't glue it back on, just held it in place.

 

yep thats what im doing as well, its just being held down by the socket clamp. i admit its a bit finicky to get it just right, but once u got it, theres no need to take off the ihs anymore (you can just repaste with the socket clamp in place).
as for the surface contacts, you can either use thin stripes of electrical tape or something as mundane as nail polish as long as its electrically not conductant and protects the contacts from potentially leaking clu, its fine. oh, and it should also be heat resistant and non-corrosive, of course i myself am using liquid conformal coating:



Sent from my HUAWEI NXT-AL10 using Tapatalk

 

Careful using that stuff, it's tough to clean up before it sets, and will leave specks of droplets all over the place.

Depending on the forumulation it will outgas for a long time, and when heated it stinks...

 

what i do to avoid droplets: spray into a small cup then dip a brush into the liquid and brush a bit onto the contacts. and since were talking about VERY small amounts actually being applied, stink is not a problem

Sent from my HUAWEI NXT-AL10 using Tapatalk

 

No, that reading is exactly how much output is leaving my UPS based on its own meter, not "from the wall" and is UPS output in watts. That is the amount of watts being pulled from the UPS by the laptop. I do not even look at the screen on the AC adapter because it does not seem to be accurate. The UPS has a USB port to send monitoring data to the OS and that is what feeds HWiNFO64.

 

Coffee Lake S will be launching alongside the brand new Z370 chipset which is known as CFL-PCH (Coffee Lake PCH). The platform will feature the LGA 1151 V2 socket so while it will have the same number of pins as the current LGA 1151 socket, it seems like Intel is moving to ditch support of the Coffee Lake CPUs on older LGA 1151 socketed motherboards such as Z270 series. It is stated that Intel will clearly be differentiating the socket with the use of “V2” moniker. The new Intel Core i5 will be the old i7 The processor lineup will launch in the third quarter with a release focused around August at Gamescom 2017.

 

great
let me get ready to sell this km1

 

OMG!! It's over!!

The time of the Double-D LGA 1151 U-P-G-R-A-Y-E-D-D is over...

Spoiler: Idiocracy...

 

It would have been quite special if Intel let us be on the same old socket. Oh'well.
And here we go...
Intel 10nm CPUs To Launch in 2018 – Feature Support on Z390 Chipset
Finally, it’s mentioned that 10nm CPUs will be arriving to the consumer family on a new chipset known as Z390 which Intel is actively working on for Q1 2018

 

wccftech

Let's remember, they're still expecting a 1080Ti and think a 980 Notebook is a 980M

okay.

How in the cow did you manage to draw 762W with two 250W GPUs and a CPU maxing at 100W xD

 

This came from a Eurocom news release about their Sky X9E3:

"The Sky X9E3 is designed to support up to 780W of power, therefore, it is capable of supporting Intel’s next generation of Skylake-X and Kaby Lake-X multiple core processors, currently under development. The Kaby Lake-X platform will provide six, eight, or ten core processors, allowing increased performance over current four-core Kaby Lake processors."

I don't know how accurate it is but they seem to think the Z270 platform will handle the new processors.

 

HAHA, man i wish my memory could hold that much. The desktops 1080Ti's memory can do around +725, before it starts to throw errors.

 

LOL!

 

My question, How are they going to ****ing cool that ****.

Please pardon my profane language, but it feel like a pipe dream or one where there will be a lot of issues. I for one an not going to buy another Clevo, until they fix their heat sink tolerances and QC.

 

Double the thickness, 4 fans and grills all they way on the rear back. Maybe make it 18. Aka P980DM-G for the i9- series And of course dual 780W

 

So the new i9 intel series are becoming a reality?

I really don't want to let myself be pulled in this, but with render times sometimes above 10 - 20 hours, I should probably consider getting a work server for renders.

BTW, I have noticed a ton of news about a new Prema vBIOS coming along. How is that going? Only beta right now, no?

 

Not in laptops

 

We don't know that the watts output shown from the GPU sensors is actually even accurate. @Prema and I already know that the mods we are working on are producing some erroneous results. They are a work in process. In Windows 7 the memory clocks read differently than they do in W10, and GPU-Z reads different than NVIDIA Inspector or HWiNFO64 on certain things.

I would be more inclined to believe what the UPS says. It has always been accurate and it does not rely on the computer to read sensors correctly. I think Brother @Johnksss@iBUYPOWER would also agree that the UPS is our most reliable source of information on this measurement.

So, on that basis I am going with what the UPS tells us as the reading that we can rely on for accuracy. The GPUs may be pulling a LOT MORE than what the drivers and GPU sensors are reporting to HWiNFO64. We already also know that the shunt mod alters the reporting of information, so I'm thinking we can take what the GPU sensors tell us with a grain of salt.

 

Have you measured max power draw with same clocks but with single graphics? Then measured the difference to see what a single card can draw? In pure watts from your ups.

 

I have not. Let me do that now.