*** Official Clevo X170KM-G/Sager NP9672M Owner's Lounge ***

Task manager shows also the ram frequency.

 

hey guys got a question , what's up with the speaker quality of this machine compared to the p870 series , man I am seeing a very horrible compared to the older versions in sound quality of internal speakers....any pointers , and i understand the speakers are on the bottom but still .....

 

P870 series actually had decent speakers for a laptop compared to many other models and brands. I usually played WoW without headphones as I found the thermals/noise on the quieter side and the speakers more than passable.

 

I was under the impression that the speakers for the X170 were supposed to be better than the P870. This laptop has a very large subwoofer for a laptop.

 

Make sure all the sound drivers and apps are installed.

 

Maybe I'm used to only having crap laptops prior to this one and my standards are low, but I feel like the speakers on this laptop go quite hard and the quality seems pretty good for a laptop. The other day I was browsing through the Steam Store for games not knowing my volume was all the way up and a video started playing. I nearly fell off my chair.

The issue might really be specific to your unit. Hopefully it's a driver/software issue.

I just uploaded a video to give people an idea of what it sounds like:

I recorded it quickly with a phone so it's obviously by no means an actual test whatsoever; video is intended to be used as a point of reference. Maybe record your own device at max volumes and see if it sounds vastly different?

 

Are you running your laptop in Performance or Entertainment mode? Mine seems stable in entertainment mode but not performance mode.

 

In Performance mode, my notebook turns off during the benchmarks, I don't know if for high temperature or energy consumption

 

Mine turns off in High performance mode too during streaming and game play. However it has to be an EC trip of either CPU thermals or Power. I can bench on mine fine it just happens now when running a twitch stream in NVEC and gaming at the same time.

Sent from my SNE-LX1 using Tapatalk

 

gaming and rendering = performance mode

But two days ago, its turn off i entertainment mode when i'm gaiming.

 

Do you have the overclocking menu turned on? Mine is most stable with it off

 

I have the same experience. OC off = stable; at the moment its on
Today the same ****... But with no load. Only windows idle => turned off without any reason

 

Are you running 2 sticks or 4?

 

At the moment 2 sticks. In the past i run 4 sticks (4x16gb 3200mhz) but this doesn't work with the newest bios version.

 

To everyone here experiencing system shutdowns: how well are you monitoring temps for CPU and SSDs? Is the CPU undervolted? I don't see these factors being discussed much.

Reasons I'm asking:

1. Without good heatsink attachment, my SSD used to climb up to an eye watering 96c+ just on regular gaming or medium loads. Applying heatsink reduced that to an absolute max of 70c but usually less.

Have you all run Crystalmark/ATTO and watched SSD temps on a longer benchmark? (Key word here is LONGER. Not a 1-3 minute benchmark). Just pop open something like HWinfo sensors monitor and let it run side by side with benchmark and actually watch it while it benches.

2. Do you have a CPU undervolt in performance mode or is the offset at 0? Even if you think you didn't set the undervolt, in older versions of the BIOS and CC, the undervolt was automatic. Maybe it snuck up on you and you never noticed it lol.

CPU lottery notwithstanding, the most I've been able to undervolt stably is -80mv on stock clocks and I consider that to be a pretty lucky result. Someone else on this forum posted a while back that upon extensive testing of the 11th Gen CPUs there was a VERY wide variation in terms of how far the CPUs could undervolt stably. Some couldn't even go below -30mv without crashing. The post might have been specific to to the 11900k though, can't recall exactly. It was posted to this forum not too long ago.

After much tweaking, I managed to reduce max burst speed temps of the CPU to the 80c range absolute max when booting up a game or other heavy app, but the average temps usually stay in the 60c-70c range during gaming with a max of 80c occasionally. I haven't had a single system shutdown outside of bad undervolting parameters early on, when I first got the laptop.

---------------

I think having and providing this type of information as a collective could help everyone with their own issues, by drawing comparisons and isolating collective issues that affect everyone and not just single users and their personal configuration.

To that note, the only thing we know thus far as a collective, without a shadow of doubt, absolutely 0% doubt, is that the system in it's current state has RAM issues, especially is you plug in 4 sticks and especially at 3200mhz. We don't know exactly what causes it but it's probably something that needs to be fixed at a BIOS/EC level. If it's a hardware problem and the system is just flawed that way then well, that's a bigger problem.

This is my own personal opinion, but if you're still operating on 4 sticks of 3200MHz RAM (maybe even just 4 sticks regardless of speeds is a problem, idk) and you're having what you THINK are RAM unrelated problems, you should reconsider. You still have the elephant sitting in the room; you have to get it to move before you can try to patch up the holes in the walls and floors.

 

@Alex Zander

Here three logs from hwinfo with Samsung 980 Pro, Samsung 970 Evo Plus & Samsung 980 while during a bench test with Samsung Magican.

And also a log from a Intel XTU while during a 5min stress test.

All benchs are done in entertainment mode.

 

@Alex Zander I just done a Test of the SSD's myself with all 3 at once doing 10GB Read and Writes with temps hitting max 60-63c, Also i don't have a CPU underclock as my CPU is locked down.
The only thing in common is i am on 4 Sticks of RAM and i still have shutdown issues as well.

However the issue that is either BIOS related or Hardware, but we can't fully tell if its the former as the BIOS is locked down tighter then a nun.

 

Those all seem normal to me. Sure the CPU runs a little hot at above 90c but that shouldn't shut down the system.

Did you try running Memtest86 after slimming down to 2 sticks of RAM to see if it fires off errors?

Edit: if you’re not getting Memtest86 errors I’d personally try moving onto running prime95 to see if for some reason that doesn’t pass. I don’t think it wouldn’t pass running at all defaults from entertainment mode on CC but who knows… in that past I’ve been able to pass cpu benchmarks on XTU and other software but then prime95 fails anyway

 

There's a user on here that has claimed they have access to an unlocked bios. Sometimes nuns open up, what can I say xD. (No offence to anyone religious here!!)

I will not claim that I have received a copy of said bios, and neither will I claim that I verified that it indeed opened up, but if you're feeling experimental and you do want to poke around an unlocked bios and are on the factory bios the system came with (i.e version "1") please feel free to look back into the forum a little bit and see if they're feeling generous.

 

Lol, Due to the ToS with PCSpecialist flashing the BIOS without permission or even changing the CPU invalidates the warrenty.

So I couldn't use it anyway, also the firmware on mine is also customized for PCS to I think as it shows their logo and bios version has PCS at the end.

However knowing the NVRAM vars helps however unable to use them because its still classed as modified bios and not sure you can even parse them anyways via efi shell.

Just a shame there is no key combo though like MSI's bios.

But for my shutdowns its either Temps or Power as Memtest on ram is fine. Will do a retest tomorrow whilst logging at 2000ms with HWInfo.


Sent from my SNE-LX1 using Tapatalk

 

Oh wow okay, got it. That's a no-go then.

But wait then, what BIOS version are you running on? Has PCS pushed out bios updates since release or are you still on the March/April release bios?

 

PCS after asking a few times (Best to call them over using ticket system) I got given

X1KMTPCS.704 and X170KM-G.02

Aka 1.07.04 for BIOS and 1.07.02 EC



Sent from my SNE-LX1 using Tapatalk

 

This is the primary reason I moved away from Samsung NVMe drives to SK Hynix models in my laptops. Much more energy efficient and run much cooler.

The variance is pretty substantial not just for undervolting but the raw pull and heat produced per chip. So far, the variance I've seen running pure auto stock (48x multi, 51x single) is 165w up to 210w running Cinebench 23 as a baseline plugging in the same 11900k class chips into a fixed motherboard set to auto with controlled ambient temps and same cooling (360w AIO). Some run very leaky and hot to boot. The beauty of true DTR laptops is you can shop around for an optimal chip. For example, you could pony up the extra money and pick up a chip from Silicon Lottery that will run cooler with lower Vcore requirements.

I absolutely agree. Maybe the best baseline for anyone having problems is to make sure you're only running two sticks, double check you have zero undervolt applied and running everything completely stock and see if you still have any problems. As you suggest, also keep HWinfo/HWmonitor running logging all temps/voltages so if you do crash, you can get an idea of your current conditions when it occurred and see if it happens again when those power/thermal thresholds are hit.

 

Hi to everyone,

My Clevo X170KM has the following config:

CPU i9 11000k
GPU Nvidia 3080 165w
RAM 32GB (16x2) Hyperx at 3200mhz
SSD M2 1TB Samsung 980pro

In this period i'm going to try several undervolt/overclok set up and at the moment i have:

GPU +160mhz clock and +1500mhz memory
CPU undervolt -140mv

and the system is stable. In Game as Battlefield 5 in 4k at ultra setting without DLSS and without RT i have an avarege of 95-105 fps dipending of the map (multiplayer 64pg)

In Unigine Heaven benchmark at resolution 1080p 8xAA, quality Ultra and tessellation extreme i got 4583 score with min fps 37 and max fps 392.

What do you think about the result? what score do you have in the same benckmark test condition?

 

What heatsink are you using? 3 out of 4 SSDs, all Samsung, are running at 55-62 degrees on average. I have 1 that is running about 88 degrees.

I also tried putting in my other 2 sticks last night with no success. I have 2 G.Skill 32GB in the back that run stable in entertainment mode but are kind of flakey in performance mode. Adding 2 more sticks under the keyboard causes the unit to not boot. I also have the overclocking menu turned off and have reset the BIOS.

 

So i have been running some tests.

Using Kombustor v4 with CPU Burner enabled at the same time in high performance mode worked fine without issues. Temps looked nice 85c max peak with 67c after PL2 finished.
Then i ran AIDA64 Stress test with everything enabled except local disks within seconds of hitting start i got an error about threading and the system shutdown.

Looking at the HWInfo log temps where fine, so the issue seems to be something RAM Related.

I have pinned it down to RAM fully now, if i swap the sticks over the system is fine and Memtest Passes no issue, as soon as 4 sticks are in hello shutdown if in performance mode.

So it looks like its a BIOS / EC issue which needs Clevo to sort it.

Edit: its also reproducible too. If you have Shutdown issues open up HWInfo to check temps and log then Open AIDA64 do a stress test with All options enabled expect Local Discs, then click start it should then start but fail soon after with an thread error. After this the system will simply freeze and shutdown.

 

Have you tried running Cinebench R20 and R23 to make sure the CPU undervolt is at least somewhat stable?
Give them a run with Hwmonitor or HWinfo running to see what your temps, wattage and effective clocks are running.

 

Thanks for doing that, that's really helpful information. Anyone having shutdown issues could try these steps to see if they can reproduce the freeze and shutdown for their own system.

And yes, it looks like the problem is out of our hands - needs to be fixed at a higher level.

 

Nothing fancy, just this guy over here: https://www.amazon.com/gp/product/B0857G3CHQ/ref=ppx_yo_dt_b_asin_title_o08_s00?ie=UTF8&psc=1

Was originally considering this: https://www.amazon.com/icepc-DIY-Gr...FJTXNV/ref=psdc_2998409011_t3_B07XD8M43P?th=1 but idk if I like the rubber bands concept

 

Were the sticks sold as a set of 4? Have you checked the sticks are the same series of chips?

 

The ram sticks are matched pairs down to month and year. Both sets of 2x 8GB Kits have also been tested on Memtest and both passed.
The issue is down to the fact something is going funky in the BIOS/EC. The fact others are having issues with Sticks even in 2 Stick configs but at 3200mhz points me to timings.

The fact that i have ran the same test setup with both sets in a two slot setup no issue, also with 3 slots no issue but with 4th slot i have an issue.
If i had another ram kit be it in 2 x or 4x i would test them too. But i don't.

I can't even test the another CPU as i don't have one extra to test with and can't do that without PCS's permission else i break ToS for Warranty, which is why i am updating them on my findings should have some news from them tomorrow though as i was on the phone to them today.

 

I've tried today so much to reproduce the turn off.

3ds max
AutoCAD
CS:GO
Prime95
FurMark
AS SSD Benchmark
VMmware Workstation

Thats the tools. I tried high cpu and gpu load, high ram load, all in different combination at the same time. All with different settings (OC in bios on/off, entertainment and power mode) but nothing. Complete stable and no problems.

What the ****...

I will try to run hwinfo in autostart and log every uptime of the notebook.

 

Try using AIDA64 system stress test and select everything other then local disks that's what causes mine to power down every time.





Sent from my SNE-LX1 using Tapatalk

 

Thanks for all the numbers a few weeks back, it's true this doesn't kick up the fans on light loads like a thinner bodied model might and the particular noise when they kick on is lower pitched even at the same volumes. Definitely sticking to these bigger bodied machines for the foreseeable future. Performance is great so far as expected for heftier tasks too.

So far about the only thing I feel like I'm missing from my older machine is the speakers. The onboard speakers on this are okay but I feel like I got really spoiled from my old Clevo P870DMX/Sager NP987X. That thing's speakers really put a lot of other machines I've used to shame, this one included..

 

Now i'm running Cinebench R23 and actually the CPU is stable only with undervolt -130mv. So i kept this value.

What i notice is during the benchmarck the CPU clock speed is stable at 4,6Ghz at 85 degree for fist 2 min, but after it drop up to 4,2Ghz for all rest of the benchmark.

Is it normally?

Also during the game as battlefield 5 at 1080p (while it run about 180-200fps) the CPU temperature is about 85-90 degree but keeping the clock at 4.6ghz without drop.

 

Might be that intel thermal velocity boost is enabled or that’s just how cinebench behaves - some part of the benchmark is less cpu intensive so it downclocks. Not sure

 

CB20 and CB23 are pretty intensive and will usually root out thermal and power constraints.

Back off of your undervolt in -10mV increments and rerun it and see not only your temps rising but more importantly if it holds its clocks.

Your multithreaded clocks should be topping out at 4.8 all core not 4.6 too.

 

CB20 and CB23 will definitely stay CPU intensive as long as you can meet the thermal and power requirements so it is most likely a power limit issue as his temps are topping out at 85c.

 

Hi everyone.

Due to the instructions received from the manufacturer of a TB3 peripheral (Universal Audio apollo x6 - audio interface) I need to somehow find a way to keep CPU's speed constant. The exact speed doesn't matter so much as long as it's higher than 4.3GHz on all cores.
The problem with this audio interface is that it produces pops & crackles even during basic audio playing with almost every change in the CPU's speed. Of course, the same happens when I'm using it's ASIO drivers while working in a DAW.

The EU-based Clevo reseller (Dream Machines) where this machine came from told me that they're no longer Prema BIOS distributors, and after going through this thread and other forums I understood that this means there are little to no chances to have my issues fixed once Prema releases a BIOS mod for this model.

So far I've tried with the CPU overclocking app in CCC and with Throttlestop. My success was extremely limited, but I'm also lacking the knowledge to get the most out of these apps.


Can anyone guide me for how could I get the CPU in my laptop running at a constant speed on all cores?


This is the configuration:
Barebone: Clevo X170KM-G; RTX 3080, 300Hz display (from an official Dream Machines reseller located in Romania)
CPU: i9-10900K
RAM: 2x32GB HyperX Impact (kit)
Internal Storage: Samsung SSDs (970 series)
I assembled the laptop using the stock cooling pads and Thermal Grizzly Kryonaut (carefully applied) on the CPU and GPU.


Thank you!

 

I checked.. the clock was going down due to the power limit set up at 130w. Now i put 170w and stay constantly at 4.6ghz.In Intel ETU i have to enable overclok thermal velocity? Because by default is off.

Moreover, why the clock maximum is always 4.6ghz?

 

If you want to disable this behavior and force both components to run at full throttle, you must disable dynamic boost.

To do this, open up Device Manager. From there, go to Software Devices, then select Nvidia Platform Controller And Framework and disable that device. Now your CPU power draw should not be lowered when running an application that stresses both the CPU and GPU.

You could disable dynamic boost from the Nvidia control panel previously, but it seems that option is now gone. The only way I know of to disable dynamic boost now is by disabling that software device I mentioned.

 

I also have thunderbolt 3 peripherals, Video & audio capture cards.
I didn't have any problem as you described, I capture multiple video with embedded audio in realtime.
What the Apollo manufacturer told you seems odd... I'm using TB3 products many years never heard about such an issue, I believe it's a driver issue.
did you try to lower the buffer at the driver app?
Also here is an updated TB3 driver from intel: https://www.necacom.net/index.php/intel/thunderbolt/12829-intel-thunderbolt-3-dch-driver-1-41-1193-0
** have you tried using LatencyMon to analyze possible causes of buffer underruns?? (launch it after a clean restart without any appa running) https://www.resplendence.com/latencymon
*** You can also upgrade to XMGs latest Bios which contains official options to disable Intel Turbo Boost & Intel Speedshift

 

Have you unlocked the windows power plans and tried the ultimate performance level?

 

So, i bought hwinfo pro and create a task. Everytime when i start the pc, hwinfo start a sensor logging into a folder. I hope we can see useful parameters, when it turned off again.

 

Make sure you logging at 2000ms though bud.

Sent from my SNE-LX1 using Tapatalk

 

You can also use GenericLogViewer from same developer making Hwinfo.

 

So, i'm running Cinebench r23 and get 14917 score with my i9 11900k.

By default the all cores run at 4.6ghz so i overclock i little bit up to 4.8ghz using for all 48x.

During the test CPU consumption was constantly about 200-210w and temperature about 95-99 degree.

Is it normaly have this temperature?

Also during battlefield 5 the cpu temperature stay about 90-95degree with 190-200 fps ultra setting (without rt and without dlss)

 

Those temperatures are the same temperature I was getting at that level of power draw, so in my opinion I'd say they're normal.

 

@Alex Zander is spot on. That's about the norm for most average specimens of 11900k at stock to draw ~195-205w at 4.8ghz all cores. You are drawing a little more because your CPUs are running hotter. It isn't uncommon for laptops running desktop CPUs to run under stock spec just to keep it cool and not throttle or always running red lined.

If you manage to snag a low leakage / cooler running / less power hungry 11900k you can put a significant dent in that draw and heat. This is when finding a pre-tested (or at least with a killer V/F curve for at least some advantage) comes into play or just go back to below stock spec at 4.6 all cores (or lower) to get it to not thermal or power throttle and keep temps somewhat sane or if you're looking to dial in a specific fan curve so you don't need to go max fans.

 

JTC taking Origin's version of the X170KM-G for a spin: