*** Official Clevo X170SM-G/Sager NP9670M Owner's Lounge ***

Two of them are dead by the looks of things.

 

And the CPU is most probably dead.

 

I kept the CPU with me in the hope of buying a temporary Z490 board for cheap and keep doing my office work in the absence of my laptop while it was undergoing warranty work at HIDevolution but found a very cheap 8700K + Asus motherboard deal that I simply couldn't refuse. I still have to test that CPU so really can't say anything. But from the looks of it, nothing looks burned or out of the ordinary and these chips are quite tough so I personally believe that the CPU will be OK. Plus, there are internal safety features built into these processors to protect them from sudden voltage surges.

 

https://www.eteknix.com/nvidia-gefo...-leaks/?utm_source=dlvr.it&utm_medium=twitter

Looks like mobile cards might be on the way.

 

I wonder how low the tdp will be compared to the desktop variant. Picture seems like the real deal as a qualification sample gpu die has already leaked, not long to go!

 

Lower GDDR6 memory speed, lower boost clocks and lower TGP vs dekstop card. Yeah, all boxes ticked, LOOL

 

They're definitely gonna have to cut it down since I believe the desktop version is 325W?

So they'd have to cut it down to around 250W at least. Looks like it MIGHT be using a desktop die, so i'm gonna assume same CUDA/Tensor/RTX cores, lower speed. Might end up performing around Desktop 2080 level as a complete ballpark guess.

 

I believe the desktop 3070 is rated for 220 watts. I'm hoping it'll only be limited to 200 watts like the RTX 2080 mobile Max-P was.

Most likely that won't be the case though. Nvidia can't let the mobile 3070 outperform the mobile 3080, so they will probably have the power cap be even lower, probably around 150 watts.

 

That would neuter their 3080 sales but I’d be very happy with that. I have some insider sources that state the 70series 115w will move up to 140w in some cases.

 

The question is, how much and what does it do to performance? The 3080 and 3090 are pretty pushed from the factory.

 

Is there anything known about X170SM-G and freezing issues? My X170SM-G finally arrived 2 days ago but it occasionally freezes. The image on the monitor gets frozen, keyboard and mouse are not reacting any more. The only option is to turn off the laptop. Does anybody have experiences with that? I already tried to re-install Windows 10 and double-checked that all drivers and the latest versions are installed. The laptop was delivered with the latest BIOS installed. Everything is running at stock (no modifications like undervolting were performed). I also run Microsoft's Memory Tester program with no issues detected.
The freezing is relatively "rare" with occurring only about once a day.

 

Check in your BIOS if you have an option called Intel SGX (which is the Intel Software Guard Extension). I had the same issue you described a couple of years ago on my Clevo P870DM3 and disabling that stopped all kinds of stuttering.

 

I experienced the same issue for many hours repeatedly until this happened. I had the SGX disabled in BIOS which could have been the only other potential culprit.
http://forum.notebookreview.com/thr...m-owners-lounge.831618/page-240#post-11049722

 

Thanks for the rapid responses and for the advice. I deactivated Intel SGX now and will report back if it helped.

Btw, I also that people recommend with similar issues to set "pci express link state power management" to "off" in the power saving settings. Do you think this could also help?

 

If anyone experiences the situation where theirs runs too hot, try this:

• Remove the three screws that have springs that tighten down the heat sink to the CPU
• Borrow three of the long bottom cover screws and put them where those used to be
• Check temps
• Buy 3 more screws for the bottom cover (or even better, buy stainless steel panhead screws to replace all of your laptop screws)

 

Do you recommend a specific part number? I have been thinking of doing the same to my machine once it comes back after the fix. Will probably use much harder springs this time or just put 3 washers before putting them back in with the same springs.

 

Oh this is interesting. I may have to try this myself.

 

Went and checked the nvidia site and you are correct. 220W for the 3070. Would be nice if it's only 200W limited since this machine has the power bricks for it. I guess it depends on the thermal capacity of our cooling? Though my GPU rarely goes over 55C.

 

You have the 2070 super which only runs at 115W, that's how yours is so "cold".

 

No, just 2.5*10mm stainless steel panhead screws. The 2.5*8mm screws in the bottom cover are just barely long enough. The 10mm will allow for washers.

I think the GPU screws may be 2mm threads. Not sure. But, I would leave the stock screws on the GPU or be extra careful about exerting too much pressure on the vapor chamber. The CPU can handle it.

http://www.laptopscrews.com/M2.5x10.htm

 

Seems like I have everything in stock except for the laptop. I will have to wait for Hidevolution.

 

So unfortunately I experienced just another freeze of my laptop while having Intel SGX and PCI express link state power management disabled. I guess there must be another problem. I checked the "reliability monitor" and at the time of freeze it only says "Windows was not properly shut down", because after the freeze the computer was not reacting and I had to turn it off. But I saw some other critical events names "Windows Audio Device Graph Isolation" stopped working. Faulting Application path: C:\Windows\System32\audiodg.exe and Fault Module name: MBAPO264.dll

Do you think this could be the cause of the freeze? Unfortunately the time of the freeze and the error with the audio device do not correspond. But I think I could at least try to re-install the driver?

Do you guys have any other idea what I could check?

 

I don't like this AT ALL! What are your cpu temps? Did you repad your system? Mine was doing very similar things before cpu power mosfets burned out.

 

I did not touch any padding or anything else. I left it as assembled by the reseller. CPU Temps of my 10900K:
Temperatures idle: 38-40
Temperatures 3DMARK Timespy: 84C - 92C (range between lowest and highest core), most of them are between 87 and 92.
Room temperature: 22C

Here are the scores:
Graphics Score 10 460
CPU Score 12 503

Details Graphics card during run: RTX 2080 SUPER (Notebook)
Clock frequency
1,905 MHz (1,560 MHz)
Average clock frequency
1,809 MHz
Memory clock frequency
1,750 MHz (1,750 MHz)
Average memory clock frequency
1,750 MHz
Average temperature
62 °C
Driver version
27.21.14.5655

 

I'd take an image backup and pop on a fresh install and minimal drivers to test if diagnosing myself. Have a word with your reseller.

 

Hopefully this isn't a flaw on this laptop, I believe another person on another thread has a similar issue.

 

Update: It seems that DHL has been keeping the laptop on hold for the last 4 days due to their suspicion that it is a laptop so it must have a battery inside it as well (even though I told the driver explicitly that the battery is NOT connected to the laptop and is packed in a separate box inside). What an epic waste of time and unnecessary frustration! The machine was supposed to reach HIDevolution today but I'm still stuck with it. What a shame!

 

Same thing happened to me with a rare GT83 I bought from my supplier. It had the 8th gen 8850h cpu and was a barebones. The parcel was stopped in Germany by dhl and 3 weeks later I received an empty box!! With a letter saying the laptop has been seized due to having a battery in it.

It didn’t have a signature on it either and the seller (my supplier) never received the laptop either making me think it’s a scam in which employees steal your stuff. Phone number on the letter went absolutely nowhere either.
Dhl is a waste of time I’d never use it. UPS is the best for worldwide shipping!

I wish you luck brother I hope you don’t loose that laptop in transit.

 

The shipping was arranged by HIDevolution as it was a warranty claim so technically, this is all on them. If something happens to the laptop then I will get a replacement machine since I didn't lose my machine. Personally I'd always stick to FedEx for international shipment.

 

DHL needs the package to be labeled a certain way for packages with lithium batteries. We emailed DHL the necessary information and it should be taken care of soon. There was only a 1 business day delay (not including the weekend).

 

Usually not, international parcel law and handler regs can get messy. Normally I'd use the main number and escalating that way.

 

do your fans rattle a bit when they start spinning?

 

Yep and the cpu side fan still rattles at a low rpm but it’s not annoying to be honest

 

Msot likely the Nvidia driver, I was having the same issue for a bit till I rolled back my driver to 451.67

You could also try this hotfix driver that just rolled out that fixes the bug thats been around for a bit here:

https://www.guru3d.com/files-details/geforce-456-71-hotfix-driver-download.html

Haven't tried it myself, as my current drivers are stable.

 

Yes, at low and medium CPU temp I hear a bit of a rattling sound. It's not very loud, though, and therefore not annoying. When the fans start going very fast at higher CPU temps, I don't hear the rattling sound any more.

Thanks for the advice! I will try this. At the moment I am trying if the update to the latest Control Centre fixes the issue.

 

Which version of the Control Center are you using? I believe there is a 3.02 version floating around out there.

 

I installed 3.06 which was posted by an XMG representative in the XMG reddit as temporary download link. I think on the official XMG drivers download page they have 3.02 listed.

Btw, speaking of it, I was wondering if you guys could help me out regarding Clevo Control Centre installation. When I uninstalled Clevo Control Centre I uninstalled "Clevo Control Centre package" and "Cleco Control Centre app" via "Add or Remove Programs". When I install 3.06 or 3.02 from scratch now, I only get "Control Centre package" back, but I can't find "Control Centre app" on the computer, which is the GUI. Do I need to install that separately somehow? I repeated uninstalling and re-installing the Control Centre now multiple times, but can't figure it out.

 

Yah, I would just try the nvidia driver (451.67).

I just wipe the control center off my computer through uninstaller then installl the control center through its installer. It should auto download the Windows apps after install.

 

Remember the fans have self cleaning functionality. They start spin the opposite way before they stop up and change back to correct direction. I expect this is one of the reasons you hear some noise when the fans start spinning up.

 

I forgot all about this laptop due distractions in the desktop arena. Are they planning to offer 3080 MXM?

 

Hi,

I think I came closer to find out what causes the freezes of my laptop. I think it is actually not caused by drivers / software.

• When I operate the laptop with 4 x 32 GB or 2 x 32 GB of Kingston HyperX Impact DDR4 3200, CL20, 1.2V, then I get the freezes (graphics freezing, sound is suddenly off, mouse and keyboard unresponsive). Then I have to turn off the computer via the power button.
• If I put in the single 8GB module of Corsair 2666MHz which came from the reseller, then I didn't experience any freezes for 2 days.
• I thought the Kingston HyperX modules might be faulty and bought 4 new modules. Once I placed them in the Laptop, after 15-30 mins there was the first freeze. After the first freeze, the freezes got more and more frequent and at some point the computer couldn't properly start any more because it crashed repeatedly during the start of Windows. But I doubt that all of those new Kingston modules can be faulty.

What do you think could be the problem?

1. Are the Kingston HyperX modules not compatible with X170SM-G? They have a JEDEC profile running at 3200HZ and CL20 and if I remember correctly it was mentioned in this thread that those sticks should be the fastest compatible sticks for that laptop. There is even one reseller in Germany (https://clevo-computer.com/) that cells this laptop model with those sticks. I think there were also 1-2 people in this forum running those sticks in their laptop?
2. Could the problem be with other components of the Laptops, e.g. could some of the memory slots be faulty?

What would you guys recommend to do? I am considering sending the Laptop to my reseller (XMG) and let them install the slower Samsung 3200MHz CL22 2x32GB sticks which they offer and let them test - even though I actually wanted to run the Kingston memory.

 

It might be the timings, but let's start with the basics. Have you tried running memtest86 or memtest86+ for an iteration or two to check to see if the DIMM modules themselves are OK?

 

I've performed a test with the "Windows Memory Diagnostic" tool with the first set of Kingston Memory Sticks. There were no issues found. I will try memtest86 now and report back!

PS: What do you mean regarding the timings? I thought the timings cannot be changed with the current X170SM-G BIOS. So I assume the sticks run as defined by the JEDEC profile?

 

I tried to run MemTest86 on the new 4 Kingston memory sticks.
1. First try: The test did run until Test 4 (or maybe 5?) and then 4-5 red errors showed up. I didn't have time to note anything down, as the Laptop aborted the test and restarted.
2. Second test: The test did run until Test 2 and then suddenly the laptop restarted
3. Third test: The test did run until Test 2 and MEmTest86 froze

Now I probably should try to test each memory stick individually?

 

Yep, at this point test individually to certify each stick.
Then take a good, known stick and test each memory slot.

MemTest86 will routinely find issues where many other programs will not. Whenever I get a new system/ram I let it run for ~12hrs to make sure everything is working as it should.

 

I honestly don't know what the Clevo X170S (or its BIOS) is expecting, and was just questioning if whatever it is using either the first XMP Profile it can match or using the JEDEC settings are compatible with the system. Something like CPU-Z within Windows would provide more info on what those settings may be.

However, since you're encountering errors with some of the memory configurations within memtest86, I think trying to resolve the memory (and getting clean memtest86 test runs) is the first thing to investigate and resolve.

 

You can also try different slot locations say 1 and 4.

 

Ok, here some interesting results:

1. I've installed only 2 Kingston 32GB memory sticks in the slots below the keyboard. I've performed memtest86 v8.4 for 4hours 37min and completed 2 passes. There were 0 errors detected.
2. Then I placed the same two memory sticks at the bottom memory slots. I performed memtest86 again and during test number 4 in the first pass, the computer freezes and re-starts itself. I tried it twice and the same thing happens.

Does this mean the bottom memory slots could be faulty?

 

Yes, that could very well be. Time to reach out to your vendor.

 

Bottom are the default. If they fail something odd is going on. Swap in the other mem sticks you have and see.