*** XMG Apex 15 with Ryzen 3950X / Clevo NH58AF1 Owner's Lounge ***

Just tried that, it worked! I don't know why exactly, but thanks so so much!!

Sent from my Pixel 4 using Tapatalk

 

Let's the machine discharge and clear any potentially floating pins or built up charge.

 

Hi guys!
Had a strange keyboard&power button freeze/disconnect while playing. Became stuck in the key I was pressing. Had to kill the power completely, removing battery and the power supply cable for it to reboot. After that it was working without any kind of issue. Seems like EC problem but never experienced it before since I use it most of the time with external keyboard. I remember someone reporting a similar issue, any toughts or solutions for this? Appreciated

Bios .06
EC .02

 

I've encountered the same issue too a couple of times, but it's infrequent enough that I havn't cared too much about it. Would be nice to know why it happens/how to fix it though.

 

The same experience here...

Sent from my SM-F907B using Tapatalk

 

See my above post, there is the physical side, also it forces a memory retrain and clears any cached settings.

 

The above issues appears to be related temperature. It only seems to happen when temperatures exceed 100c for me anyway which no longer happens.

 

Well, that was not my case. Temp's rarely exceed 65°c both gpu/cpu on my case. But... I can't be sure about Vrm, chipset or even vRam as I've seen some weird artifacting. Now for precaution I run the fans a bit higher to see if it relates.

 

I can second this; I always undervolt my cpu and basically never exceed 65C either. My guess is a buggy FnKey driver personally.

 

I actually don't use and don't have any clevo software installed. No clue what's the issue here. But i'm inclined to software/firmware and not hardware.

 

I noticed that my power settings (which were tuned for extremely aggressive parking and higher promotion thresholds for normal usage) were negatively affecting my Cinebench scores. The image is with a 3.850 Ghz clock, before and after resetting my power settings to Windows Balanced (but still leaving PBO disabled). The laptop reaches about 86C, so there's still a little headroom (maybe about 8750 with a 3.9Ghz).

 

It's a balance towards different traits rather than better usually with power settings.

 

I actually sent my system in for the exact keyboard issues that people have posted. I'm still waiting to get it back but now I'm not feeling hopeful that it's going to be fixed.

It feels like a software issue. When it occurs I can't turn the system off via the power button. External keyboard will work fine. If I go to the fan control it complains about a critical service no longer running. My CPU/GPU temps aren't being read either but I've been keeping and eye on my thermals. They were below 80c before the issue.

 

I think theres too many of us with identical symptoms for it to be a hardware issue. Are you able to use an external keyboard when the bug triggers?

 

Unless it's a batch issue I suppose but service errors are suspicious.

 

Yes I can use an external keyboard when the bug triggers. It gives me the opportunity to troublehsoot but I really haven't tried a lot. My image was fairly clean and all my drivers were up to date so that's why I was quick to think it was hardware. I tried some steps though. I removed the Obsidian fan control application and installed the Clevo Control center back on the PC. Rebooted, removed the control center, rebooted, installed the Obsidian fan control. I really don't know that either of those apps are the cause but it is odd that whatever is crashing is impacting obsidians ability to read the thermals.

 

Well done review my friend. I did the like and subscribe thing as well. You have a great relaxed and natural presentation and give the right info at the right pace. I get so nervous and feel uncomfortable on video, I do better in giving live presentations but get nervous for those too. I guess it's having more experience with them that gives more comfort. Did you know that public speaking is one of the two greatest fears of the human species? The other is physical confrontation, and many experienced, skilled fighters have told me (I trained as a hobbyist in a professional Muay Thai gym while I lived in Thailand for 10 years) that the worst thing about a fight is the interview you have if you win.

 

That is awesome, I love the Linux OS but am forced to use Windows due to software requirements of my bachelor's degree program. That's actually why I had to upgrade to the Clevo NH58AC, to have the processing power to get through Excel when applying array formulas to spreadsheets with hundreds of thousands of rows of data and running Tableau on similar data sets.

I notice little difference between 32GB RAM and recently adding 16GB more, from my old, unused laptop, so I must not be putting it to use the way that you are. I don't do much compiling in Windows, my work is almost all database right now with some business web application development work using Visual Studio (ASP and C#.)

I did learn that there is a significant difference in performance between installing VStudio on my OS SSD (A Crucial P NVMe 1TB) compared to installation on a second NVME 2TB drive made by Mushkin. I couldn't believe it but benchmarks when the computer arrived and I had added the second SSD showed something like 4 or 5 times better performance from the Mushkin. That 2TB NVMe was the same price as the Crucial 1TB, and I always thought of Crucial as a good quality lower cost alternative to Samsung and OCZ.

To be fair, the OS runs from the Crucial and may have an effect on it's performance, especially when running VStudio on that drive while the OS also accesses it. In any event, running the IDE and programming enviroment on the second drive loads applications in 5 seconds or less, while on the C drive it takes up to 10 seconds to begin.

 

Which drive is installed to the Gen3 x4 slot? Whichever one is should perform the best given that it is directly connected to the CPU PCIe lanes. The other slot is connected to Gen2 lanes via the B450 chipset. You might consider swapping them if the drive with higher throughput is in the slower slot.

 

It's 2x gen 2 lanes too iirc? Certainly want the cpu connected one.

 

One of my important computing tasks takes up to 4GB ram per thread(!) so sometimes I disable hyperthreading when running it, and even then I cap out. Other memory hogs include running multiple virtual machines at once, which runs much smoother with more RAM.

Check that your ssd is in the pcie 3 slot - mine came from aftershock in the pice 2.

 

That's a decent chunk of ram for a thread lol. A portable VM box is the use case I saw for it.

 

I'm pretty happy with how my unit handles VR, I get 72C on the GPU (2070) and 63C on the CPU (3900 set to 3.3GHz at 0.975V) and a steady 80 frames per second on the highest settings in HL Alyx with a Rift S. For less graphics intensive VR apps I normally limit the GPU's clock speed to 1500 MHz and get even lower temps with pretty low fan noise. It's nice to have a portable VR setup like this

 

I tried everything about that but even without these apps in execution (well I am not sure about services) and that still happened. The issue happened once/twice every two weeks and in the end I decided to remove the battery completely. At least, this sped up the restart. I tried even flashing the most recent BIOS available without having it resolved. I tried latest XMG BIOS and at the moment it has not yet appened again, but it's only two weeks and it may be a coincidence. Note that this voids your warranty - but I have no possibility to send it back for "repair" - thing that I doubt is even possible in case of a software/firmware problem.

By the way, I have a PC Specialist version, and PC Specialist Service was absolutely terrible. They negated the issue happening with PC Specialist firmware, (well I was the living proof this was not true), attributed the issue to Windows asking for a clean install (sic, and everything is running perfectly except for the keyboard issue) , I pointed out them this thread saying there were other people across several vendors having the same issue but they negated again and when in all honesty I said that I tried to use a non-PCS BIOS for solving the issue they banned me from the forum for "wasting their time" when my intention was to point out to them a potential flaw and not certainly wasting time. But evidently it was untolerable fo them to admit publicly there was something fishy about this keyboard issue. Absolute ********. I wonder if other Clevo vendors behave this way, because this is a very serious problem making me wonder if I would ever go again for a custom laptop from their subvendors (from PC Specialist I will not buy anything more for sure - and this was my second laptop from them).

 

Maybe water coolers?
https://pl.aliexpress.com/item/4001...earchweb0_0,searchweb201602_,searchweb201603_

 

Maybe, but maybe an ITX system

 

Have you tried undervolting with afterburner and curve editor? You may be able to run 1500MHz@700mv which should decrease temps/noise quite a bit compared to underclock which probably has vol5age running around 800-900mv@1500MHz.

 

I hope Clevo all8ws for Ryzen 5000 support. We know the powerusage is on pair with 3000-series and the regular 65W limit should be a good option here aswell. Seems like they use the same nodes as on the XT-series which makes it possible to run 2-300MHz higher at same voltage compared to the usual yield of non-XT-parts. Perhaps 4GHz is doable at 1V?

Has anyone tried ram OC above 3200? Is it possible through ryzen master? I would guess that relaxing voltage sensitive timings like the 3 primaries, tRAS, tRC and tRFC would make it possible to run ram faster than 3200. From what I have seen good ICs can easily do 3200cl20, some do cl18. 3600@cl22 should be possible if you can change it i RM.

 

looks like we might be screwed

[email protected]​

6:10 AM (2 hours ago)


to me

Dear ----,
Because the new line up CPUs all have wattage over 65W, they will not be included in NH57ADS's compatibility list.
Thank you for your understanding,

 

Well, that is not true for the 5600X, this CPU has 65W TDP. And, BTW, the 3950X used in many current configurations has 105W TDP but it is running in ECO Mode, that is limited at 65W (88W in reality). So it's the easy excuse for not doing anything and force you to buy a new model...

 

Yeah, that's exactly what I am doing! I am limiting the gpu with afterburner set to 1500MHz at all voltages.

 

That is concerning, but hopefully this rep is just misinformed. The 5000 series already have the 5600x and there are likely to be other 65w models in the near future. There are also no indications that eco-mode is no longer a thing or if a 65w 5900 is planned.

I think we will just have to wait for some official feedback from resellers, XMG, Sager, etc before we really know what the plan is.

 

I hope so for those looking for an upgrade, personally I am more than satisfied with my 3700X.

 

Yes they are run in TDP down mode. The chips would need to support that.

 

A 5600 vanilla pr 5700X should be viable when they arrive since they are probably 65W

 

Anyone have this laptop with R5 3600? If so what temperatures are you on CPU?

 

Let me know you make out with the XMG BIOS. I have the Prema BIOS. I guess we will see how HiDevolution handles this keyboard issue. I linked this forum to them so hopefully they understand that it's more than just my system. They are going to load a game and see if they can replicate the issue. Hopefully they can but it's pretty intermittent for me. At least a few times a week though and always in a game.

 

FWIW both times I encountered this bug was while playing Shadows of Mordor (which relies heavily on space-bar, ctrl and shift keys). Agree its pretty rare (maybe once every 10-15 hours playing the game).

I looked at my windows events (event viewer -> system -> filter current log -> critical+error); nothing immediately stood out, but I have many error ACPI events of the form:

: The embedded controller (EC) did not respond within the specified timeout period. This may indicate that there is an error in the EC hardware or firmware or that the BIOS is accessing the EC incorrectly. You should check with your computer manufacturer for an upgraded BIOS. In some situations, this error may cause the computer to function incorrectly.

Hopefully it doesn't require a BIOS upgrade - the golden rule is to never patch what (almost) isn't broken.

 

@up i have Dream Machines, but it is the same like XMG Apex 15 with AMD. I have no problems with the keyboard. I play for several hours a day. My bios is 1.07.07TDES

 

That can vary from chip to chip and machine to machine.

 

All depends on the MB manufacturers... http://forum.notebookreview.com/thr...ers-welcome-too.810490/page-883#post-11051942



And Notebooks is in an own league regarding compatibility/upgradeability.

 

So far it did not happen again (fingers crossed)

In my case it happened only when Playing World of Tanks and World of Warships. So possibly this is tied to specific software conflicts with the keyboard controller.

The XMG BIOS version I installed should be based on the same code (as it is the 1.07.07RTR1_r2).

In any case, it could be tied to specific hardware/software combination. I.e. I have RTX2070, two SSD hard drives (1 Samsung 970 Evo 1 Terabyte, 1 PCS 1 Tbyte) and 32 Gbytes of RAM, and Killer NIC.
And, as told above, it happened only when playing WoT/WoWs (but these are also my most played games).

 

I was playing Black Desert Online. I'm not sure if this matters either but the game relies on shift and space for quite a bit of things. At fist I thought it could have been some of the accessibility options in Windows like sticky keys so I disabled them. The problem seems lower level than that though. Driver or BIOS probably.

My system config is a bit different: x3950 cpu, rtx2070, 64GB memory, 2 hard drives (1 Samsung 970 pro and 1 western digital ssd, intel wireless, win 10 pro.)

 

I need help. Do you have users with 64GB (25600) RAM? (CPU: 3950X ) I use two 32GB 25600 (default 3200), but if I lower the TM5 test to 2666 it will only last 22 minutes and turn off.


Help

 

That has been my rule of thumb, I'm still running a BIOS and EC that XMG put out in May/June (first one with virtualization support enabled). I'll have to check the specific version of each and see if there are meaning quality of life updates available yet... or at least any that are worth the risk of doing the updates

Just to clarify, Asus released a statement on this saying that the support rep was misinformed. Asus clarified that they are planning to provide 4xx boards with BIOS updates for Zen 3 once AMD makes this available in January.

 

My reseller released an update yesterday EC firmware version 1.07.04. Check the XMG page, maybe you also have.

 

Unfortunately this evening it happened again, after almost 20 days. So the frequency seems lowered but the issue is still there.
Well, it is not that it is unusable, but of course it's annoying.

 

That's unfortunate. What CPU are you running? My system is on the way back to me. They reseated the keyboard and reinstalled the keyboard driver. I'll let you know how I make out but I've already reinstalled the kb driver so I'm not hopeful which is going to be a pretty big problem for me.

There's some company representatives here for this model. Have your support people seen this issue before?

 

I'm not aware of anything in particular.

 

I am running a 3700X.