*** Official Clevo P870DM2/P870DM3 (Sager NP9873/NP9872) Owner's Lounge! - The Phoenix 2 is here! **

I have been using galistan (LM) for several years in temperature monitoring thermowells (where we need to know the temperature of a chemical solution down to the tenth of a degree C). It has replaced metallic mercury that we have used for +50 years (industrial process control). It has not evaporated (dried up) or had a problem with oxidation as the surface area is very small. Inside of an IHS heat spreader that is adhered down to the CPU carrier PCB it is not going to give you problems.

You can get sloppy with the material, "mo is not betta" and you want to minimize the size of the gap you are trying to close up with any thermal material. It should be metal on metal contact with the LM acting more like a wetting agent between the two surfaces to deal with irregularities and parallelism. If you pull apart a heat sink and find a half a mm of goo between the items you should correct that problem first. Compared to the thermal conductivity of copper, nickel, silver (or aluminum.. not with LM!) any compound is going to act like an insulating layer. LM just happens to be the highest in thermal conductivity so heat can flow across the gap between the two metals.

Even in thermowells with four wire platinum RTD temperature probes we do not rely upon galistan as the primary way to conduct heat or cold to the probe. Those probes are spring loaded so there is a constant force pushing the probe down against the bottom of the thermowell. The galistan just increases the surface area in that little dimple at the bottom of the well so the probe reacts faster to temperature changes. The same should be true with a heat sink to a CPU or a IHS to the die package, there is already a mechanical attachment and the LM is just enhancing that by increasing the surface area in contact.

 

i was actually thinking of going gallinstan instead of CLU, seeing as the stuff is pracitcally the same but at a tenth of the brand name price

 

@FreshGuy6204

Ive got a cheap Skylake chip that @jaybee83 was referring to that would work for that purpose. I only bought it for the IHS. PM me if you're interested since it's no good to me. It still has an IHS it's just been delidded and polished.

 

This is how these posts sound to me:


Note: Link shamelessly stolen from forums OT sections credit @katalin_2003

 

That is an assault on the brain to an engineer because they throw in a couple of actual words

 

Much appreciated but I've bought a 6700 non-K already...
I'm staying away from delidding forever now.

 

http://rockitcool.myshopify.com/

 

I was @Prema 'd today. I love the enter screen after the power unplug it said "free your mind or world or something like that". Anyways. Had to leave her behind to go back to work. Can't wait to check over it tonight when I get home and see all the new options. Does this bios have the battery option insem to remember in the older Phoenix that allows only certain percentage charge of the battery so it doesn't stay fully charged at all times?

Yay! That dang clevo flash screen is finally gone!

The EVOC screen looks bada$$ - really polished the machine off.

 

Psh, take it in with you and tinker with it during lunch break

 

I normally would.. had a crashed sql server and corrupt database I've been dealing for a week now, ugh...

 

Ouch, a week? No tape or replication backup?

 

The database was 230'ish gb. I was brought in literally a few days before this happened to take over from previous fired IT company. I was putting scope/RFP and plan in place as this happened. Luckily Ontrack was able to rebuild he SQL db from the failed raid array. But data reconstruction for the third part interface is about to be underway. So a couple more days.
Previous company had a three drive raid 5 dell poweredge r320 with SATA instead of SAS drives and 16gb of ram running 5 VM's one of them being the sql server with that db with 190'sih employees accessing it. So now I have a dual 16 core server with 128gb ram, 2xraid 5 sets with 5 1.8tb 10k SAS drives in each. Veem setup with replication, external backup and that replicated to my cloud server I have setup at my office.

 

I've had more issues with my Sky X9E2 computer since I had Eurocom flash the Prema bios. Now I can't get more that 0.773 volts to the processor. It's at a multiplier of 19. I installed @Mr. Fox bios settings in the prema bios and still nothing. I had to ask Eurocom support for the new CCC version. Then ask them to send me the updated video drivers for 7700k. What would be causing these issues? Is there a new firm ware mod I need, EC, what ever? Any help would be appreciated.

 

Just set Windows power plan to performance.

 

I love how Prema always solves an issue worth of multiple paragraphs in one sms sentence

 

Thanks @Prema, it already was set to high performance. I have no clue what's going on. I appreciate your help

 

My control center isn't going up the scale. It's stopping Half way to 4.2. With 0.772 volts

 

And XTU is lit up with current limit throttling

 

Get rid of CCC & XTU
Load BIOS defaults
Shutdown
unplug AC Adapter
Plug it back
reboot

 

You're the best @Prema, I'll give that a try and get back to you. I have to get you a consulting fee. Thanks again

 

@Prema No Idea how you know this s**t but it worked. You're the best. I owe ya big. Thank you. Should I forget the CCC and XTU install all together? Or do I need CCC for just the fans. Thanks again.
Jon

 

@Jon Webb you can install CCC from @John@OBSIDIAN-PC OBSIDIAN Tools ( https://www.obsidian-pc.com/files/OBSIDIAN TOOL/OBSIDIAN_TOOLS-1.01.07.zip) lets you install CCC without XTU.

Note to future people reading this: at the time of writing, the link for the latest version is in @John@OBSIDIAN-PC's signature.

 

Oh it can be painful sometimes to see what has been put in place, though sometimes the companies can be at fault too. I had one expecting a pure HDD v7000 to respond within the listed manufacturers average response time of a single HDD and wanted a report each time levels raised higher. I tried to get them to get a tiered SSD cache setup recommended for them since response time was important but they would not listen.

 

I don't understand how so many people have issues with Intel XTU? This is mind boggling to me.

 

Also:

Edit the OEM.INI file in the Clevo Control Center's directory before you install it:

Change SupportXTUFanTable=1 to SupportXTUFanTable=0

@Jon Webb

 

That is freaking hilarious! I recognized every item in the youtube clip but as the dialogue went on I just could not keep myself from laughing.

 

Well if you exclude the issues it caused with CCC and fan controls in some BIOS/EC versions, I never liked it because I could not figure out how to make it start automatically, so I would constantly have to go back in there to make sure my settings were still active. That seems like fairly obvious requirement for software nowadays but it never had it for some reason.

Without auto-start everytime I had a crash it would go back to stock settings, sometimes it would be a week before I noticed (sometimes I may not have been sitting at the PC when it crashed and didn't know it did) and was troubleshooting why it was running hot or not at the clocks I had set, only to find out it reset at some point. Wasted lots of hours on it and I think it is a terrible design for software. At least Throttlestop (which seems overly complicated to me) can be auto-started making sure the settings that I want are active.

I just don't have time for software like that. I want to boot up to my desktop with all my stuff running and be ready for whatever task I throw at my machine.

 

Well considering Intel XTU is not installed with CCC and you have to install it separately is task number 1

You have a 6700K which can run without you touching anything in the bios up to 4.6 ghz. (Unless your EC is broken)

Intel XTU (Not CCC-XTU) can make changes that Throttle stop can't when you lose os bios access.

You would only use the Intel version to make changes when throttle stop can't make them.

Throttle stop works best with it's profiles and independent voltage controls, but when the os locks out part of Throttle stop, guess what. As long as XTU is not greyed out it can recover and allow Throttle stop to function again.

Your settings revert back because they are not set right in the bios. If you set something moderate then on a crash it will reboot those settings (Most of the time). (Set something that is suspect and the bios will reset. Usually due to watchdog detecting a hanging system.) If you put XTU in the startup folder it will autoload a set of settings for you, but TS is still better in that respect.

So that brings me back to why so many problems? If you are talking about CCC-XTU, then that's fine because that is not what i'm talking about.

 

I was speaking in general terms, I've used XTU on many machines, most time on a BGA system to reduce heat. I used to use it for undervolt but not anymore as it's too unpredictable.

I didn't know some of the things you mentioned tho, +rep for that

 

@Johnksss@iBUYPOWER, IIRC, installing Intel XTU would cause random weirdness with the fan tables. For example fans would ramp up to 100% even though the system was cool as a cucumber while sitting idle.

@Phoenix, do you recall who traced this behavior back to XTU?

 

Really, which version of XTU?

 

That is a really good question. I want to say the last one in which this was NOT a problem ended in *.50, and the ones after introduced the problem.

Sorry I'm not much use for this. A search on XTU and fan tables in the forum might shed some more light.

 

I think in some cases the options are too complex and spread out for some users and that some of it is the options selected.

 

Hmmm... My recollection was that the options were either not changed are slightly modified after an upgrade of XTU to whatever version it was and then the fan craziness started up.

Surely someone who actually encountered this can retell their story.

 

@D2 Ultima Here is a link to a video just after it crashed while gaming . Would it be possible to tell which component ( CPU, GPU, Motherboard or PSU) could be the problem based on this?

 

Either an OS or a motherboard issue, I'm almost certain. CPU/GPU crashing will turn the system off nearly immediately. Any audio freezing/stuttering like that happens for a short time while the system is crashing, not after it's already crashed and is sitting there with your screen off and your keyboard backlight clearly on and the system frozen solid.

I would RMA that *IMMEDIATELY*.

 

what are you BIOS settings, power limits, voltage, CPU clock speeds?

Can you try setting your RAM to XMP Profile 2 in the 2nd overclocking menu under memory then save / exit. it will take a while to boot as the BIOS is adjusting the new memory timings so don't be scsared

 

Been sitting here for an hour waiting, I'm guessing thy aren't going to do my Prema update tonight afterall. I wish they would have at least called to cancel so I wouldn't waste my evening waiting for something that may never come...

 

Too funny...your appt was between 4~5 PM Pacific Time, and he called you at about 4:50 PM...

 

Happen also to me.. during gaming screen become white and after restarting i was with the same problems.. blinking screen and audio problems.. after a while disappear.. However i solve to change the thermal pads on gpu's and don't happen anymore..
motherboard overheating maybe...

 

He did finally call. It said 4-4:30pm pacific, which is 6pm here, but he did finally call like you said. I am a happy camper now! Thanks HID!

 

awesome man! lemme know if you need any help with the BIOS settings

 

Will do, thanks! I almost took you up on it, was scratching my head trying to figure out why benchmarks were capped at 120fps. I checked gsync, vsync, sli, rebooted a dozen times.

I had a framecap in RTSS at 120 that I forgot about...

It looks like I lost about 600 pts in FS, I'll need to figure out what's going on. 1100 graphics points and a few hundred cpu and combined

 

I tried with both stock Clevo BIOS 1.05.02 ( no overclock whatsoever / voltage offset @ a mere 4GHz) and the same with the latest Prema BIOS, the output is the same. It is excellent for gaming when it works though, couldn't be more amazed by the sheer performance ( I previously owned a Alienware M17x R3). @Phoenix I will try XMP Profile 2, even though it had crashed with the Default Memory Profile @2400 MHz.

 

@D2 Ultima Actually, it used to happen like this only for a short time and system did restart after a few seconds, but then I disabled Automatic Restart on system failure in Windows http://imgur.com/a/PXys6 hoping to get a Blue Screen to get more info about this issue. After this change, Windows doesn't restart automatically and I do a forced restart everytime.

One thing that I did notice in 90% of the cases was, it crashes when CPU/GPU utilization goes from 0 to 100% in a short period of time. For instance, it crashes in the first few minutes after starting a game or when I minimize and then maximize a game after some time. Either way, I guess an RMA is the only way ahead since it looks more and more like an hardware issue .

 

Do you have those crashes in all games?
Did you try to use another NVIDIA driver and if so which ones?

 

I also had these random crashes when gaming, though they seemed to have disappeared after I got the Prema Mod Bios. Fingers Crossed.

 

Yes, it pretty much happened atleast once in every one of the 4 games I have played so far. I have tried 378.49, 376.33, 376.60.

 

If you are using a 3k or 4k screen set Windows to 100% scaling (disabled), play in native resolution see if that fixes it.

 

Also, test the memory to make sure that is not the culprit. It has happen to me more times than I care to count.

Side note.

I got my hands on some of this and now need to test it against everything else we have been using. No AC only max fans with cooling pads and without cooling pads. Thanks.

http://forum.kingpincooling.com/showthread.php?p=31830#post31830