All Activity

Maybe you make test

So the motherboard is dead?

I have the 170cm model as well. It’s equipped with an i9-10900K CPU, 32GB of RAM, and an RTX 3080 GPU. I haven't encountered an issue like this. The latest XMG BIOS version is installed. Your RAM slots might be faulty; either slot 3 or 4—or both—could be the problem.

you win some, you lose some: GPU cable keeping steady, still at +/-10% max. current imbalance, roughly equating to 1-1.3A, very happy with that. but now suddenly my machine kept crashing during RE9 Requiem gameplay, sudden black screen followed by a reboot, not BSOD, no warning, no freezing, nothing, just instant reboot. had a hunch so i ran TM5 and voila: the RAM started getting unstable! currently retuning the timings while keeping the speed at 7600. not sure if this is silicon degradation or rather the current heatwave. ambient temps are reaching 28-30C in my little gaming corner in the living room, thus pushing the RAM to the upper 50s during gameplay (remember: this is with RGB switched off and a 140mm fan pointing directly towards them). in any case, back to RAM tuning ugh....altho not much in the mood, crazily enjoying my gaming phase 😄 ill just do it on the side, one timing a day i guess.... 😛 at least i found the culprit quickly and can go back to stable gaming without reboots 🙂

So single channel works in slot 1 and 2. When I try single channel for slot 3 or 4. It doesn't boot. CPU socket pins, everything looks clean. Nothing is broken or missing. Used an air blower and a brush to clean the socket, still the dual channel doesn't work. Is that a motherboard issue or a CPU issue?

Hello. I'll try with another CPU. Like the 10100F. I have a none bent socket. Found one CPU locally for 35 after taxes

The processor's memory controller might be faulty, or there could be an issue with one of the RAM slots.

Can you send me a link to your backlight mod v.2 please ? I believe this Is your concept , Yes?

OK... The decapitation was painless and the thermal improvement was even greater than my other 270K Plus (which was about 15°C). Using the same BIOS settings, more than 20°C cooler in Cinebench R23 versus stock solder and IHS. The IPC uplift with 270K Plus over 14th Gen Intel and AMD's finest is pretty insane. (Captured in the CPU-Z benchmark.)

I try this with 2 slots (1 and 2) in single: 1x8GB (2666MHz) works at 2133Mhz, boots to OS 2x8GB (2666MHz) works at 2133MHz, boots to OS 16GB (3200MHz) + 8GB (2666MHz) works at 2133MHz, boots to OS 2x16GB (3200MHz) doesn't work, no boot. 1x16 (3200MHz), doesn't work, no boot. Anything in dual channel, no boot. All the RAM sticks get hot in the touch, so they are getting powered. Is that a CPU problem?

forgot to mention i had to elongate a bit on the holes on the delta to able to screw in the bolt near the fan connector.

I tryout a experiment 4 years ago with a dell m6700 cpu first avc fan then delta fan for my m6600 but i forgot to do the "see the different temps after sacrificing the m6600 cpu fan connector" before do this mod😅, what i remember at that time the fan able to go 100% fan speed at time of how high the fans can go. it able to turn on and off normally. now currently the fan hasn't gone up 100% as I now I'm using a different thermal paste gc-extreme and not the mx 4, and put thermal pad to send some of the heat from cpu to the palm rest. i haven't post about this mod cause i didn't have proper test results and its currently using optimus with p4000 cause the amd m6100 gave up so temp are different vs the time i was testing it. the cpu is i7-2760qm. if possible if the fan can fit and the holes align on your m4600 if u want to trying out the experiment on m4600 but there risk if not done right. as i wasn't using m6600 as my main laptop. i was able to solder the cables to the position where the delta fans old cables suppose to as the delta cables had color marking nearing the solder pads for which cables go to where. If u do try it make sure the fan can turn on n off/lower the fan rpm as my first test on the delta fan i had to go adjust the solder of the cable to able to turn off/sends commands to lower the rpm properly when the temps goes down enough.

You're absolutely right CO can definitely trash your memory OC but we also had this to a lesser degree even back to 6th gen Intel I can recall experiencing having to adjust my memory OC to reflect core OC. Had to strike that synergy between memory and cores but back then I focused on clock OC then would shape my memory OC around it. Since the advent of AM5 and especially AL, I tune memory first and lock it down then lock down a D2D core OC shaped around that. As for curve optimizing, I tend to test static OCs first to try and get an idea of what can be done fixed then work with CO around that. I started doing that when testing 5800x chips and found static and CO all tended to lead to the same place but CO gives you better per core dynamic performance and you can have your cake (all core OC) and eat it too (even higher per core boosting).

I learned a lot about Ryzen by chasing overclocked benchmark scores. It is goofy in some ways compared to Intel. The same in other ways. Principles are the same. Clock is king and clocks need more voltage. Ryzen starts losing performance when it is voltage starved. It clock stretches but performs worse, maybe starts dropping IPC. TM5 (as well as Karhu and perhaps other tests) is also going to error out if your Curve Optimizer undervolt is too aggressive. Don't automatically blame it on the memory or a weak IMC. I'd recommend setting Curve Optimizer offsets to 0 when checking for memory stability, if all tests pass and you set your previous Curve Optimizer offsets and start having errors, you know immediately what needs to be addressed and it's not the memory overclock. You can use the SMU debug tool to test PBO CO offsets with Core Cycler in Windows, then apply the settings in the BIOS once you find the max stable negative offset. Best to do it per core. There can be a gigantic difference between cores. Perhaps contrary to popular belief, a core that can handle a deeper negative offset is not necessarily a good one. When I look at per core SP ratings and V/F curves, the better cores are already needing less voltage and the crappier cores need a deeper negative offset to match the voltage needs of the better cores.

Really depends on what level of stability you're seeking, but I usually start with a basic 90 min run of TM5 to kick off anything as that usually covers all my bases for basic D2D use. If I find memory errors or issues gaming pop up, almost like clockwork TM5 will flag errors. Just for first tier testing, I toss in WMD and even memtest and OCCT. If I can pass all four, for me personal needs that is usually good enough. I'm not going to subject my laptops to Karhu or similar but if you want to make things spicy you can give that a whirl.

No, this is my Zephyrus G14, but the security principles are surprisingly similar. And during my first attempts to unlock the hidden menu, I was only ever able to access different parts of the hidden menu, exactly as shown in your photos.

Is this on Area-51m? The main issue is, I don't know how to bypass boot guard. It requires signed bios to boot , even if you spi flash it doesn't boot.

Hello. I bought a Clevo X170SM-G with 10900KF and RTX 3080 16GB with 32GB of RAM. It has the watercooling 2in1 heatsink. Laptop wasn't posting so I reflashed it with a 1.07.04 stock clean ME bios. Still no boot. So I had to remove the RAM from slot 2/4 and I only used one slot and it worked. The laptop only posts with single channel configuration and it runs at 2133MHz. The previous owner told me that the laptop was working perfectly fine and he didn't use it for a week and the laptop wouldn't post. I cleaned the CPU pads. There were some stains but this didn't help. Is the motherboard dead? Or is the CPU DDR controller dead? All the memory slots are working in single channel.

So far I have only tested with OCCT's 1 hour combined load test but I know that errors can pop up beyond that length. That also makes me wonder about the 7945HX3D as I was only testing 1 hour on it, and on top of that it was being tested in the winter months where the ambient temperatures on average were 5-10F cooler. Is there a guide for a better combined load test setup that the community uses or would paying for OCCT be the best option to test beyond 1 hour? I also spent a bit of time testing a tuned 5600 profile with VDDIO/VDD/VDDQ at 1.3v. No crashes or errors but memory temps do creep up over 95C in combined loads so I will probably need some kind of memory cooling. One nice thing about the MSI AMD bios is that it has a Save/Load overclocking profile function that works, so its actually easy to switch around configurations or restore them after a bios reset. Yes it was a full assembly (A-cover, camera, camera cable, display, eDP cable, L+R hinge, bezel/B-cover). Talking with another friend who replaced just a display panel, he noted that MSI was using something that looked like glue so removing an existing panel from the A-cover would most likely crack the panel in the process. Not a great thing if this is being done as an upgrade swap instead of replacing an already damaged panel. I'd like to find another full display assembly like this so I can try out a 1200p panel as well.

Could you tell me how you managed to access these menus? Maybe I can study the method as well and see if I can help with this.

I’m honestly really happy and grateful to see such talented and knowledgeable people on this forum. A big thank you to ssj92, janepa, and all the other members who generously share their knowledge, skills, and expertise with the community. It’s truly appreciated.

I'm sorry, but that message offering help was primarily intended for the user ssj92 as a small gesture of support after hundreds and hundreds of wasted hours and failed attempts. The kind of work that most of us can't even begin to imagine how difficult and invaluable it is. He knows exactly what I mean. Once again, I'm really sorry. It's my bad.

With modded Bios - Before put the 1070 mxm gpu You must to choose an SG mode into bios then save the settings then turn off the laptop then put the 1070 gpu into the master mxm slot

I saw this and it made me really happy.👍🏼

User ssj92 is also working on the same project and has made some progress. If you are familiar with this BIOS, perhaps you could work together with ssj92 and finally unlock this damn BIOS. Unfortunately, I’m not familiar with Aptio or its editing environment. However, if you need any files from my side, I can provide them and help however I can.