Tanaz

I've had an RX 6950 XT for almost 2 years now and ever since I got the GPU I've been having black screen/ grey screen crashes ranging from a few times a week to sometimes a few times a day. Well in the last few days I've had enough and I've done an enormous amount of troubleshooting and have figured some things out. First before anyone asks let me list the things I've tried to resolve my issues: Different Windows versions and windows reinstalls (both Windows 10 and 11). Disabling MPO. Since I have 4 monitors I've actually tried unplugging them all individually and using them in all possible configurations (single monitor,dual,triple etc). Undervolting the GPU. Overclocking the GPU Both setting the powerslider to less than 100% and to more than 100%. Using my GPUs secondary quiet vBios. Uninstalling drivers with DDU (with ethernet unplugged so windows doesn't install drivers on its own) and testing at least 4 different versions of both Adrenaline and Pro drivers. I'm using 3 different PCI-E cables. Tried 2 different PSUs (my current one is a 1000W Seasonic Platinum rated) Disabled XMP/PBO and had my entire system running stock. There are actually more things I've done but I can't even remember to list them all at the moment. Now for the big revalation of my troubleshooting: Grey screen crashes with blue lines and black screen crashes are the same crash! How did I figure it out? Well I'm lucky to have 4 (actually 5) completely different monitors and I realized I'm getting a grey screen crash only on my OLED 1440P ROG monitor and on my LG C2 tv. I think the reason it's a grey screen with blue lines and not a black screen is because of either the 10 bit color depth or HDR being enabled. That's why you see so many people with Samsung G7,G8 and G9 monitors complaining about grey screens with vertical blue lines and they think their issues are unique and it's because of Samsung but actually they're getting the same black screen crash most people are getting only in a grey color since those are the most widespread HDR/10 bit monitors. How did I confirm this? There's a game called Predecessor that's literally the only game in which I can 100% of the time replicate the grey screen/black screen crashing. This game for some reason can almost immediately crash my GPU and is the perfect testing ground for troubleshooting. To be clear I'm still getting crashes in regular destkop use and in other games, they're just not happening as often and are not as easy to replicate. So when I unplugged my OLED and was testing my 1440p VA 8 bit monitor as my main to see if my PC would crash it would still crash, only this time it would be a black screen crash due to the lack of either HDR or 10 bit ( not exactly sure). Okay.. so far so good but there's still the issue of..what the issue actually is. And to be frank.. I don't know. Seems like those issues really accelerated after 22.5.1 since AMD started changing the way their DX11 inner workings function. I'm sorry I don't have the technical knowledge to fully explain, but that's when all sorts of dx11 acceleration issues in browsers popped up and when people started experiencing more and more crashes. In my amateur opinion, having tested multiple different monitor, driver and OS configurations I think that either there's a hardware flaw in these GPUs (which would be weird given the fact that even people with 7000 series cards have grey screen/ black screen crashes) or there's something fundamentally wrong with the drivers. So what are the solutions? Linux is one of them, as I've rarely if ever seen people complain about AMD GPUs working properly there but given the fact that the user base is extremely small and most popular games don't even run because of anti-cheat issues (eazy anti cheat not working in linux) I can't say I'm confident these GPU bugs don't exist there either. I've spend so many hours in trial and error and at this point the only solution for me personally is to go back to Nvidia. I feel extremely annoyed because I'm going to have to buy an overpriced 4080 right at the end of the GPU cycle but I simply can not have my PC crash at random times just because it's using graphics acceleration (something GPUS ARE MADE FOR). The other day it even crashed in VS Code while I was in the middle of an important project. And lastly for the people that are bound to comment with "bro I've had an AMD GPU for 45 years and have not experienced a single issue" - well congrats to you! That doesn't invalidate all the bad experiences people have had and are continuing to have on a daily basis. But at least if someone from AMD reads this the one important note they should take from it is that grey screen crashes = black screen crashes.

I don't know about combining intel with AMD drivers because Intel drivers are still quite new ( not to throw shade at the Intel team that've been progressing at a rapid pace) but I do know that AMD and Nvidia GPUs work well together so you could go with something like a second hand 3050 and get good results.

So on my personal website I have a circular gradient background that uses mouse tracking to move dynamically. I did this by using the mousemove event listener like this: "use client"; import { useState, useEffect } from "react"; export const useMousePosition = () => { const [mousePosition, setMousePosition] = useState({ x: 0, y: 0 }); const updateMousePosition = (e) => { setMousePosition({ x: e.clientX, y: e.clientY }); }; useEffect(() => { window.addEventListener("mousemove", updateMousePosition); return () => { window.removeEventListener("mousemove", updateMousePosition); }; }, []); return mousePosition; }; and then adding in-line styling to my jsx page that use the x and y position of the mouse to move the background circle gradient. I want to use this gradient as a "flashlight" effect to illuminate the text of my H1 and for it's shadow to move dynamically based on where the mouse cursor is. The issue is that while this mouse tracking function works perfectly with my gradient it does NOT with the text shadow because the shadow css property accepts both positive and negative X and Y coordinate values that calculates the position of the shadow based on the element its attached to while the mouse event listener uses only positive values. How do I reuse this function ( or make another one) that translates the numbers to negative based on my h1 positioning or is there a CSS property I might be missing? I know this is hard to explain so I will provide a short video: https://streamable.com/a20kge By using style={{ textShadow: `-${x / 20}px -${y / 20}px 14px rgba(0, 0, 0, 0.61)`, }} Negative X and Y values (divided by 20 so the shadow can stay closer to the h1 element) I can get the text to move up and to the left but not down and to the right. It's difficult to explain better, hopefully someone will understand what I'm trying to achieve.

Now I'm having another issue. I'm trying to animate each word to fade-in one by one but the moment "by Atanas Atanasov" appears the h1 jumps up for a second. Any ideas what might be causing this? Here's a video. And here's the modified CSS code: .gridContainer { display: grid; width: 100vw; height: 100vh; grid-template-columns: repeat(20, minmax(15px 15px)); grid-template-rows: repeat(20, minmax(15px 15px)); } .h1Container { margin: auto; grid-row: 8/16; grid-column: 8/12; font-size: 55px; overflow: hidden; visibility: hidden; } .byAtanas { grid-row: 14; grid-column: 10/12; justify-self: end; /* Align the paragraph to the end of its grid cell */ font-size: 20px; text-align: right; animation: fadeIn 2s; animation-delay: 6s; animation-fill-mode: forwards; } .scrollForProjects { grid-row: 18; grid-column: 1/20; margin: auto; font-size: 2em; } .scrollLink { scroll-behavior: smooth !important; } .scrollButtonContainer { grid-row: 20; grid-column: 1/20; } .h1span1 { animation: fadeIn 2s; animation-fill-mode: forwards; } .h1span2 { animation: fadeIn 2s; animation-delay: 2s; animation-fill-mode: forwards; } .h1span3 { animation: fadeIn 2s; animation-delay: 4s; animation-fill-mode: forwards; } @keyframes fadeIn { 0% { opacity: 0; width: 0; } 10% { opacity: 0.1; width: 100%; visibility: visible; } 100% { opacity: 1; visibility: visible; } } edit: I animated "width:0;" accidentally and forgot to delete it. Fixed.

So they need to be treated as a single grid item? Do I just do something like <h1><br><p></p></h1> inside a single div that's treated as a grid item? edit: Yep that's exactly what it is. Put both the h1 and the p inside a single div to be treated as a single grid item and didn't even need a <br>. It worked. I spent hours on this. I'm livid to say the least. Thanks a lot!

I'm using NextJS so the code might seem a bit weird for some but what I'm having issues with is a very basic web layout with a big H1 in the middle and my name right below it aligned with the right edge of the H1. The issue is that my grid needs to be 100vh and 100vw sized and not a fixed size because I'm making a "snap scroll" type of website that would scroll an entire section every time a button is pressed / the scroll wheel is used. So whenever the viewport gets resized the positioning of my H1 and <p> relative to eachother changes. I know it sounds hard to visualize so I have made a short video demonstrating the issue (I've also enabled firefox grid view for better visualization of the issue). As you can see "by Atanas Atanasov" moves relative to the H1 whenever I resize the viewport and I need it to be aligned with h1's right edge. Here's my section component's code (ignore the button and "scroll down for projects" paragraph, those are not relevant to the issue): const Section1 = () => { return ( <section id="section1"> <div className={styles.gridContainer}> <div className={styles.h1Container}> <h1 className={ralewayH1.className}>Design.Develop.Deliver.</h1> </div> <div className={styles.byAtanas}> <p className={ralewayNormal.className}>by Atanas Atanasov</p> </div> <div className={styles.scrollForProjects}> <p className={ralewayNormal.className}>Scroll for projects</p> </div> <div className={styles.scrollbuttonContainer}> <Link className={styles.scrollLink} href="#section2"> Click </Link> </div> </div> </section> ); }; export default Section1; And here's my CSS configuration that quite frankly looks bonkers after all my attempts to fix the issue (latest idea was to split the entire viewport into a 20x20 grid and position the items precisely but that obviously didn't work): .gridContainer { display: grid; width: 100vw; height: 100vh; grid-template-columns: repeat(20, minmax(15px 15px)); grid-template-rows: repeat(20, minmax(15px 15px)); } .h1Container { margin: auto; grid-row: 8/16; grid-column: 8/12; font-size: 5em; overflow: hidden; } .byAtanas { grid-row: 14; grid-column: 10/12; justify-self: end; /* Align the paragraph to the end of its grid cell */ font-size: 2em; } .scrollForProjects { grid-row: 18; grid-column: 1/20; margin: auto; font-size: 2em; } .scrollLink { scroll-behavior: smooth !important; } This is an incredibly basic issue yet my brain's been fried trying to fix it. Any help would be appreciated . Thanks.

I've had this same issue with my 6950XT in a few games since I bought the card almost 2 years ago. A lot of people say that disabling MPO fixes their issue but it didn't fix it for me. The issue comes from the GPU for sure tho.

Update: Decided to update you all on the state of this monitor. ASUS actually listened (after months of complaining) and released Firmware version 105 (their fourth attempt of trying to fix the HDR color issue) and it WORKED! The colors in HDR are much closer to natural and now rival my LG C2 in terms of accuracy. I still consider more than half a year of waiting for a flagship monitor to be fixed unacceptable and would strongly consider waiting for a few months if you're looking to buy a monitor from ASUS' new QD OLED line-up because if their team is still incompetent at color calibrating and take months to fix simple issues it doesn't matter if it's WOLED or QD OLED, it would still make for a bad product experience.

My 6950XT still has issues playing back 1440p/4k videos without stuttering on my multi-monitor setup. And this has been a well-known,documented bug that supposedly got "fixed" in October of last year yet it hasn't been. DX11 video rendering is bugged in Firefox AND Chromium based browsers (reproducable in Linux as well) in certain sircumstances. However some people don't have the issue and that's the reason why it's low priority for the driver team. I think having multiple high refreshrate monitors ( I have 240hz,165hz,120hz and 120hz 4 monitor setup all monitors having different resolutions) is just a recipe for disaster but I've also heard people with NVIDIA GPUs having issues with black screens and driver time outs on such setups so I don't know what the solution is.

LG firmwares start out good and then get even better with time. Great example is my C2, firmware was good and color accuracy was fine but there were some issues with auto dimming in certain movies, and they fixed it in the latest firmware. Meanwhile ASUS somehow have managed to mess up firmwares on multiple monitors and it's not even an OLED thing, they mess up VA and IPS monitors too. Their forums are full with enraged users, it's just so sad to see how far they've fallen from grace.

These are not gen 1 OLEDs tho. They're using the same MLA WOLED tech that LG's been perfecting for 10 years. But to your point they are gen 1 "products" and that's why most of what could go wrong outside of the display itself has gone wrong - firmware, warranty, color calibration etc.

I purchased the ASUS PG27AQDM around a month ago for it to become my new main gaming monitor. I play OW,Fortnite and League as my main games but I also fire up the occasional AAA game. So I will be concise in my explanation as to why you're much better off saving your money and going with the LG variant that is 300EUR cheaper in Europe and a few hundred dollars cheaper in the USA: 1.) The brightness - this is the main reason people consider this monitor an upgrade to its LG variant. Oh, it has a heatsink and it's much brighter, well as someone who has an LG C2 TV right above it I can tell you that the brightness difference isn't nearly as big as people make it out to be. The thing about OLEDs is that they're either bright enough for you or they are simply not. It's a perception issue more than it is an actual brightness issue as nits don't tell the whole story, seeing the display in person does. So if you're going for the ASUS because of brightness , just don't. 2. HDR is broken. As in horrendously, almost unusably broken. There is a serious problem with the colors in HDR mode, everything is orange/red and it's actually almost decent when watching HDR content in HDR mode but the minute you play SDR content in HDR mode everything is bright orange/red like you're on the planet Mars. This is the 4th firmware revision, ASUS are aware of it and they still have not managed to find a person that possesses the gift of sight to properly calibrate the display. 3. The OSD/firmware itself. Speaking of firmware revisions, you can not touch the color sliders when the monitor is in HDR mode so you can't manually calibrate it even if you want to and you are stuck with the horrible orange color accuracy. Another issue with the firwmare is the fact that the "pixel move" function that's available in most OLED displays to protect them from burn-in is by far the worst I've seen on any display ever. It's so agressive even in it's light mode that it cuts off half the Windows clock, so you can imagine just how many columns of pixels you're losing. In normal OLED displays it's usually 2-3 rows and here it's more like 10. Another firmware issue is the bright red borders on every menu, which definitely can not be healthy for the display. 4. The stupid RGB Lighting. You can control the RGB only on the back of the monitor and not on the front. The bright red ROG logo in the middle of the display base can not be turned off and it turns orange when the display is in stand-by. That's right, you can't turn it off and you can't change its color. The base of the stand as well as the "projection ROG logo" also glows bright red but at least it can be turned off. So in essence the only "RGB" in the whole monitor is the logo on the back that YOU CAN NOT SEE, and everything that's facing you is this bright red garbage. 5. The ROG legendary warranty is no longer a thing (dead pixel). 5-6 years ago I bought a top of the line ROG laptop and my warranty was such that if the laptop exhibits any defects during its warranty period ASUS would not only fix it and return it to me but give me a full refund as well. That is how legendary ROG warranty used to be. Nowadays?.. not so much. My monitor came with a dead subpixel but ROG warranty is pretty much the same as LG or most manufacturers (except Dell who actually have a 0 dead pixel policy on a lot of models) and requires >=5 dead pixels. I could've returned the monitor to the retailer as I am in Europe and I have 14 days to do so so but that's not a "warranty" return,it's an online purchase law return. But point still stands that the warranty is the absolute bare minimum AND it does NOT cover burn-in either. Paying the ROG premium just doesn't give you anything other than a logo anymore. So yeah, get the LG and save your money, the ASUS has worse firmware,worse HDR, bad RGB lighting, absolute bare minimum warranty and is not bright enough to justify the price premium.

I don't know if the monitor can give you permanent eye damage but this post has given me permanent brain damage.

To me it's a case of popularity outgrowing competency. It happens, I don't think it's malicious however malicious or not the results are still the same. I don't feel more negatively about Linus/LTT because I've never had sky high expectations of them. For me they've always been that fun youtube channel that occasionally will have some of the greatest tech banger videos on YouTube. I've never expected a data-driven, accuracy focused channel so I was never disappointed when I didn't get one. In terms of the Madison situation, again it's popularity outgrowing competency. Does that mean there are not any malicious people working there? No, I do believe there are a few bad apples at the very least that need to be removed, I just don't think it was this great conspiracy cover-up scheme, it was more of an underestimation of the severity of the situation combined with not enough/wrong information being fed to management. So if that gets taken care of my sentiment would be exactly where it was a month ago - positive/neutral.

At the start of this entire drama I was 100% on GN's side and I genuinely believed that there is no agenda or a goal to tear LTT down. After watching TechTechPotato's video my perspective started to shift a little bit and after this deleted GN video it shifted further. Now don't get me wrong, 2 wrongs don't make a right, I think Linus 100% deserved to get called out but I do also have this sense that Steve wants to tear down LTT for his own personal reasons that have nothing to do with objective journalism. There I said it, and in case I'm blamed for flip-flopping - yes that is what you're supposed to do when presented with new information. If something else comes to light I might "flip-flop" again.

Now this was a good video. For me at least I can safely say I've found closure and I'm also confident that they'll do justice to Madison and remove the culprits responsible for the toxic workplace environment. Considering just how low LTT's turnover rate is I think it's safe to say that it's a case of overwhelming majority of people being good and professional with just a few bad apples inbetween.

Obligatory "I didn't watch any of the videos reacting to the drama but I read the comments" lol.

An update for those who care - I've decided not to return the monitor. Outside of the dead pixel the panel is actually flawless both in terms of uniformity and lack of dirty screen effect / color tinting. OLED panels seem to have a much higher chance for dead pixels so there's no guarantee that I'll get a better panel after I return this one, in fact chances are I'll get a worse one. I'm just going to have to fight my OCD and not stare at the pixel, which is not that difficult considering it's invisible in 50% of situations.

I don't know, maybe I'm overreacting. It's not even the entire pixel that's dead, I think it's the white subpixel because I can only see it being black on a white or blue background. On a red/green/yellow background it's not dead. So I technically have 1/4 of a dead pixel. edit: So white and blue subpixels are dead,red and green are working. The question is is half a dead pixel worth the entire hassle of repackaging, paying for delivery back to the store and getting my money back.

I meant in general not in my personal situation. When I was looking into buying my C2 TV there were dozens of threads with people with WOLEDs getting 1 dead pixel at first and then it spreading into a cluster. So my 1 dead pixel can turn into a few rather quickly. And it's not even at the edge of the screen it's in the upper right quadrant so if it starts to spread it would be mega visible.

It doesn't but I've never seen the amount of dead pixels and dead pixel clusters on normal LCD tech. If it was my TV I'd be fine with it but given that it's a monitor that I'm looking at up close it's kinda tough to get over.

Yeah I think I'll do that but not sure if I'm going to get a replacement or straight up get a Samsung G7/G8 mini-led and not worry about the whole OLED fragility.

Well ain't that just peachy.. I have a dead pixel ALREADY and it's been like 40hrs since I got the monitor. God damn it.

You really have quite the sample size. It'd be an interesting experiment indeed. I have the C2 and now the ASUS so I have a sample size of 2 WOLEDs, let's see what happens.

I agree with your sentiment. It's called a "gaming monitor". If playing games is "improper use" then perhaps they should market it as something else. I'm going to use it as I normally do and whatever happens happens. I would say I'd rely on "that ROG great warranty" but that's very much a thing of the past, seems like they've decided to become way more hostile towards consumers recently.