All Activity

If anybody cares or finds this while having your own issues. I eventually sent it into Dell. From my testing it appears to me that 4 red 1 blue has to do with the charging circuit on the motherboard. Very well could be a battery issue but I believe there is a code for a faulty battery. Could possibly be an over draw or possibly overheating issue on that circuit or the battery. Not sure. But again I think it is a mobo issue. dell… My Dell support interaction was interesting to say the least but I won’t get into it… I will say Dell does have a few very nice people working for them. Just gotta find them. Had a supervisor reach out to me which was very cool. Dell tested it and then replaced the mobo and heatsink assembly. They said it was overheating, it wasn’t when it left but 🤷🏻‍♂️ They then tested the battery and said the battery was fine. last update I got was it will be load tested for an extended period of time and should be sent back in a few days. hopefully it’s fixed but again I think those lights have to do with the mobo and hopefully fixed my issue. Could be wrong, it has happened once or twice. lol

Just get a 100% working CMOS battery, set the bios to UEFI and no LEGACY OPROM and there you go. RTX Ada working on eDP 1080p 120Hz. No Optimus, no MUX Switch. If we set to UEFI and LEGACY OPROM enabled, the GPU displays 4-5 blinking on the top left corner of the screen and then... Nothing. No boot, laptop stays on. Now I need to figure out how to disable the LEGACY OPROM. I'll probably disassemble the whole laptop, unplug the eDP LCD and force the boot to HDMI out. Legacy works with HDMI because I tried once and I can disable LEGACY OPROM so I'll obtain the eDP display back. I enabled the legacy oprom to use the shell grub...

Hello, have you decided broblema? I have the same situation

Yeah @safedisk was saying I was the only one having issues and it was either my settings or cooling were not right. Totally not true. A few others posted that any AGESA after 1.2.7.0 has been problematic for them as well. I figured it out (what needed to be changed) and saved the profiles, but I will be using BIOS flashback to downgrade both machines to older/better firmware that does not have AGESA 1.3.0.0 or 1.3.0.0a. The 2101 version had some kind of issue with Bitlocker keys not working or something like that. That does not matter to me one iota since I disable Bitlocker, but it would be a major problem for anyone that does. The Kingbank 8400 48GB kit in the Strix will not even flash the new BIOS correctly. It hangs flashing "zone 5" and soft-bricks the mobo. It can only be flashed using the BIOS flashback method, so you know that something isn't as it should be. There is no reason for the flash process to fail with BIOS defaults set, and I tested the process twice. It failed both times and required BIOS flashback recovery both times.

Nice quick guide and one thing I like (and many other hardware w/ software makers could learn) is to allow the ability to completely decouple software from the hardware or at least allow settings to be retained in the hardware and software uninstalled at a minimum. EVGA installs were minimum and never required them to remain installed with their hardware. It looks like Asus *might* be learning with their new web based approach to setting up various devices and using them. The Armory Crate backlash may have finally caught up to them. No surprise with China. They have over 4x the population size (~1.41billion vs 345million) than the US and continue to advance and grow as a first world, consumer based, high tech market of western proportions at a voracious rate with plenty of disposable income. They are a massive market that continues to grow for consumer goods and products from McDonalds to Nvidia, their citizens want it and companies want to sell it. I wouldn't be surprised if the Steam Survey continues to skew towards them. We went through the same thing with Japan ~30 years ago when exports were grossly larger than imports and they were a tech paradise buying up everything. To realign the market like the world did with Japan will take a bit more coordinated effort with China but eventually a certain level of give and take has to be established. Personally, I've been to China and loved it. I also love all my state side Chinese friends and family. I'd go back in a heartbeat to visit. -------------------------------------------------- As for the drivers, good to see it was yet another bug. My big takeaway from it all with @Mr. Fox testing is EVC2 can override the limitations placed by Nvidia even if accidentally and If they were to implement such a bogus system, all cards would operate normally, better binned / better voltage models would continue to boost better naturally and only if someone tried to implement a decently sized OC would it then cap you at the knees which is trash. Even 5070 owners were reporting being capped when OC'ing. ----------------------------------------------------- Nvidia will continue to "innovate" by offering tiers of cards and performance while restricting memory as much as possible. Based on their need to redirect memory, the 5050 9GB doesn't make sense but the return of the 3060 (GDDR6), 5060 8GB variants and even a 5070 8GB makes sense for their goals. I wouldn't be surprised to see a 5070ti 12GB suddenly appear..... consumers low man on the totem pole as always when push comes to shove.... I installed the newest beta, but the last two I installed, 2101 and 2004, were problematic for me and I actually had to bump my voltage up by .05v to get them stable-ish. Previously they would boot and perform fine at 1.40 but suddenly i needed 1.45 at 8000 (8200 rated TG sticks) to even get them to post let alone pass any tests. I ended up rolling back to 2133/6400 settings with 2101 because it just would not stabilize at 8000 and I was getting sporadic crashes in WoW which can be memory sensitive. Long time players know the dreaded "Memory could not be 'read'" error..... I did notice Asus pulled 2101 completely in favor for 2102 beta for the x870e hero so something must have been off with it. I also noticed this in the notes for 2102 beta "2.Provide additional stability margin during high-frequency DDR5 training. 3.Addresses potential boot failures and stability issues observed on certain Ryzen 9000 configurations." I'm going to run 2133/6400 for a bit on 2102 to make sure it is rock solid, then recertify at 8000 again in a few weeks.

I think you should try more just turning on the laptop. When RAM is replaced in p775 or p870 it takes a lots of reboot and turning on and off for me. (It takes so much time you start considering the status of the new components but after it starts. It looks the same like a new bios flash) FN+D is the default settings you can give it a try.

I get crashes from around 95 watts onwards in-game, vrms are all covered i used silicon paste on them instead of pads as was worried they weren't touching the copper heatsink, I've seen your post on the backlight pin modification, I will check this out further. The only way it works better is setting the core clock to lower MHz, it's stable with my maximum on 1350mhz, I'm going to tinker tonight with the settings and see what I can do with the voltage unlock. Can't afford to buy the hp version they are very expensive around £800

90W no issues. So anything above 100W = issue. Are your VRM/Coils cooled? Thermal pads touching the heatsink? You should try with the RTX 5000 HP variant and see the results. It's a 110W to 115W max TDP card and see if the TDP issue persists or not. HP is better because the GPU has the backlight pins integrated in the circuit board.

Anything eDP is related to the dGPU unlike the M6800. The M6800 can do iGPU or dGPU in eDP. The M6700 has 40 pins and 4 lines so that's good. Easy 1080p 120Hz and 1440p 120Hz. The M6800 is stuck at 1080p 60Hz.

I still seem to have an issue with power, it managed to complete a GPU test on 90 watts, any higher wattage causes crashes, so far the best thing I've been able to do is limit the core MHz and not tamper with the power. I have the screen and got a edp cable coming I'll keep the current screen on and boot the bios and throw it into PEG mode and try the other screen out, it's worth a go to see if by chance I can get it to work, should I be looking at changing it to muxless or keep it on fixed and full

I was referring to M4700/M6700 here. These systems can only use iGPU with an LVDS panel.

or u can order this: Für Himbeer Pi 5 RTC Batterie kasten Echtzeituhr halter mit Batterie/ohne Batterie Echtzeituhr Batterie modul - AliExpress 502 i needed for my old PC, much easier to handle with open, put battery in, n done... when ur system is still plugged in, even no bios data loss. and u need to check the plug size, maybe u need a different model for the m6700

uhhh, so the m6800 is a massive step forward 😄

Hi everyone! ☺️ I'm struggling with bringing my Clevo P751DM2 (i7-7700K, GTX 1070, Prema BIOS) back to life, and hope to find help here. It turns on but doesn't boot. It hasn't been my main PC for like 5 years, and has been unused and stored in the box until about 1 year ago. Here is what happened. A) ~1 year ago. I decided to re-use it as a secondary PC, took it out from the box, booted it into BIOS and checked the settings. Then, I decided to open it and re-paste the CPU and GPU. I removed the heat-sink and took the CPU out to clean it from the old thermal paste better. Suddenly, the IHS fell off. First I panicked, but after a short internet-search and some reddit threads i thought that this is nothing critical, and some people actually detach it intentionally for delidding. So I just cleaned the CPU and the IHS from the old thermal paste, applied Arctic MX4 on the naked CPU, applied the IHS on top back where it was, applied the paste on it, fixed the CPU in the socket, cleaned/re-pasted the GPU, and assembled everything. A few weeks later, I decided to start preparing it. I booted it into BIOS a couple of times. Then I changed the RAM configuration, replaced the CMOS battery (included waiting for a few minutes to reset the BIOS). At some point, it supposed to re-POST (and get back to BIOS after self-restarting several times) as RAM configuration changed, however it didn't happen. Here's the behavior that started: When turning on, the screen remained black, no BIOS, no anything. The laptop stayed powered on, with the lamps on, but with zero activity and black screen without lit. And when pressing Fn+1, the fans didn't start blasting at the maximal speed (which I thought is a very low-level thing that should work at all times as long as the laptop is powered on). After some minutes (~3-5 minutes I guess), the fans suddenly started to blast at maximal speeds, and a minute later the laptop turns off. When powering on again, the same stuff repeats. I had to stop and postpone doing anything with the laptop. B) Now. A few days ago I came back to the laptop with the aim to find out what the hell is going on, possibly fix it and start using it. I powered it on several times, with the same result as before. Tried removing all RAM and keeping just one brick, removing the battery and using only the adapter. I opened ChatGPT, described the situation, and asked what it could be. But when I mentioned that IHS fell off, ChatGPT raised a red flag, stated that this is bad and the best would be to get a new CPU instead of trying to re-apply IHS back. It also claimed that simply re-pasting the naked CPU and re-apply IHS back is not quite ok, can cause various issues, and a special high-quality thermal adhesive must be carefully applied instead. I checked other AI chats (Duck, Proton, etc.), and they didn't rise any red flags and even listed re-applying the IHS back as one of the solutions. But at least now I got a candidate that potentially zombied the laptop. Before checking the online markets, I decided to give the original CPU the last try. I opened the whole thing again, detached the IHS, re-pasted the naked CPU, applied the IHS back on top, re-pasted the IHS and fixed the CPU in the socket. The laptop didn't get back to life. However, the behavior changed: Now, when powering on, the laptop stays idle with the lamps on and screen turned off like before. A few minutes later, the fans don't start raging like 1 year ago and the laptop just reboots itself. This repeats several times. After several such repetitions, it stops rebooting and just turns off. The situation with Fn+1 remains the same - nothing happens. What I also noticed, is that the battery takes about 1 whole day to be fully charged, according to the lamp colour. I started charging it yesterday afternoon with the unknown current battery charge, the lamp has been remaining orange until the late night, and only today in the morning it was green. I decided to re-plug the charger and see what happens. I re-plugged it, the lamp still remain orange after 5 hours have passed. What the hell's happening with the charging. I'm not sure if this situation was the case 1 year ago. I think it wasn't (but i can be wrong). Does anyone know what's going on, what could be causing the issue? Could it be indeed the CPU, and is replacing it worth a try? Or is the situation worse than that? I would be thankful about any advice or help with getting it back to life.

I already did that, very hard to remove the soldered joints but was able to get a working battery.

I finally got the Apex and Strix stable with AGESA 1.3.0.0 and 1.3.0.0a but overall it wasn't worth the effort. I am not identifying any benefit whatsoever to having the newer firmware. Performance is slightly lower with the "latest and greatest" cancer... as usual. Cinebench scores are lower and old memory tuning profiles that were stable before were no longer stable. M_ORDERING changes according to @safedisk. ODT values also had to be tweaked for my systems to avoid Karhu and TM5 errors.

You can replace it for cheap with a regular cr2032 battery and some tape, see here :

See: The 7760 GPU is physically larger and will require physical mods in order to "make it fit".

Okay, thanks

Me and one guy weren't able to get the RTX 3000 from HP to work with the M17X R4 eDP. In LEGACY and in UEFI. It's a 8 beeps. However with a M6700 eDP, the RTX 3000 from HP works in both LEGACY and in UEFI. Also you have the ADLINK version of the RTX 5000 which means your card lacks the backlight pins.

MSI is working all good, was me being a fool, ive ordered the edp cable for the m17x r4 - i will investigate the screen next, doesnt arrive till the end of the month

MSI afterburner has the same power functionality as the the shell smi one.

Ok the divers 528.24 allow me to change the watts but only in power shell it runs stable on 90w and also 100w, I don't have to adjust the core clock speed now, I'll investigate how to set MSI afterburner so I don't need to do this. Looking good so far

Okay let me try. Everytime I switch the settings, it get resetted maybe because of the CMOS battery, that's why I gave up. Let me switch that.

Hello, they are impacted by the TDP bug, 40W or fluctuating between max TDP to 40/55W. So it's not worth it since they are limited by the Optimus MUX Switch. I got the M6800 to work with the modern RTX Ada and there was no TDP bugs.