GPU Artifacts when under heavy load.

[weed detector]

Hello! Since I've gotten my PC 3 years ago, there's been this weird issue that has been haunting me since I got the PC. The issue is that the GPU starts acting weird and I have to hard shut off my PC with the power button. There are three types of this artifacting. First type is where the image becomes a mess. The second type is where the image freezes up to a random solid color (depending on what image is shown). The third type is the same as the second type but also adds in white lines that look detailed a bit. I have tried a different monitor with a different cable but no luck. My power supply is a grey box LC POWER(?). It's 500w but has no PCI cable. The issue makes me think it's the graphics card because the graphics card has no place to place in a PCI cable, And if it were a lower wattage PSU like 330w Then it would be sufficient enough power for my graphics card. Any thoughts?

[weed detector]

引用自 Bad 💀 Motha：

Just use CPU I'D HWMONITOR it's much bette4. Speccy is better suited for viewing certain in depth info, not so much for temp monitoring. You want to always use an app that can record and show the current and max at the same time.

What's your full specs?
To easily share this, in Speccy click File > Publish > copy to clipboard. Paste the link here.

For RX 250 it might be a good idea to DDU wipe your drivers and then reboot and try an older AMD gpu driver. Like one from 2019 before Adrenaline 2020 came out. If you can't find them on amd.com look on guru3d.com

Operating System
Windows 10 Pro 64-bit
CPU
AMD Athlon X4 840 97 °C
Kaveri 28nm Technology
RAM
12.0GB Dual-Channel DDR3 (11-11-11-28)
Motherboard
ASUSTeK COMPUTER INC. A68HM-K (FM2+ ) 68 °C
Graphics
Generic Non-PnP Monitor (1440x900@60Hz)
1024MB ATI AMD Radeon R7 250 Series (C.P. Technology) 54 °C
Storage
465GB TOSHIBA DT01ACA050 (SATA ) 30 °C
119GB SAMSUNG MZYLF128HCHP-000L2 (SATA (SSD)) 29 °C
Optical Drives
No optical disk drives detected
Audio
Realtek High Definition Audio

[Electric Cupcake]

Wakey wakey, time to bakey.
https://forums.tomshardware.com/threads/baking-your-graphics-card-in-the-oven.794207/

[weed detector]

引用自 ZootsIdenticalTwinSisterDingo：

Wakey wakey, time to bakey.
https://forums.tomshardware.com/threads/baking-your-graphics-card-in-the-oven.794207/

I don't wanna do that, I have a bigger risk frying my card than actually fixing it. Also the artifacts freeze up my PC and don't show instantly when i boot into Windows. Vram might be dying.

[Set-115689]

Maybe look into the cpu heat?

I wouldn't buy a 710 video card.

Not sure if your system will meet minimum specs or if you can get around the min specs. If you have a good internet connection and a very high/unlimited data cap playing on geforce now might work. Maybe 1-15 gb per hour? You use their computers to run the game while you play it on your pc. The free version also has a 1 hour time limit per session. You have to log back in every hour. No daily limit. Be careful about internet data usage and your data cap or you could end up with a large internet overage bill. You might get banned from mmos for using a vpn since geforce now kind of ends up working that way.

https://www.nvidia.com/en-us/geforce-now/games/

[weed detector]

引用自 Set-115689：

Maybe look into the cpu heat?

I wouldn't buy a 710 video card.

Not sure if your system will meet minimum specs or if you can get around the min specs. If you have a good internet connection and a very high/unlimited data cap playing on geforce now might work. Maybe 1-15 gb per hour? You use their computers to run the game while you play it on your pc. The free version also has a 1 hour time limit per session. You have to log back in every hour. No daily limit. Be careful about internet data usage and your data cap or you could end up with a large internet overage bill. You might get banned from mmos for using a vpn since geforce now kind of ends up working that way.

https://www.nvidia.com/en-us/geforce-now/games/

I'm planning to play games as is, not in Geforce Now. I don't really care if I have to do weird stuff like Low Spec Gamer to get the game to launch, I want to just play games with the 710 and live with it. My budget is very low anyway and despite how much monthly I earn, it would take a huge chunk of my money even for a 710. Remember, most of my money that I have to use for like food and bills is gonna be taken out on a trash graphics card. I don't feel like hunting for used deals right now so that's why I'll get the 710. I don't think it'll be that bad since it's gonna be a GDDR5 model.

[Bad 💀 Motha]

引用自 Overheatin：

引用自 ZootsIdenticalTwinSisterDingo：

Wakey wakey, time to bakey.
https://forums.tomshardware.com/threads/baking-your-graphics-card-in-the-oven.794207/

I don't wanna do that, I have a bigger risk frying my card than actually fixing it. Also the artifacts freeze up my PC and don't show instantly when i boot into Windows. Vram might be dying.

If vram was dying it would always have artifacts even in the bios screen areas. It could be happening only after the gpu gets too hot (54*C is idle, you need to see what it hits as a max when under a full load). It could also be a poor unstable OC, even a factory OC of sorts that is now all of a sudden unstable. Or a simple driver issue. But usually with a driver issue you would generally get a proper BSOD or game crash error, crash to desktop, etc.

I'll ignore that 97*C cause there is not way that sensor is accurate. But a lockup or reboot, it is possible the cpu gets too hot while gaming also.

Contact the gpu maker and see if your Radeon 250 is still under warranty.

[weed detector]

引用自 Bad 💀 Motha：

引用自 Overheatin：

I don't wanna do that, I have a bigger risk frying my card than actually fixing it. Also the artifacts freeze up my PC and don't show instantly when i boot into Windows. Vram might be dying.

If vram was dying it would always have artifacts even in the bios screen areas. It could be happening only after the gpu gets too hot (54*C is idle, you need to see what it hits as a max when under a full load). It could also be a poor unstable OC, even a factory OC of sorts that is now all of a sudden unstable. Or a simple driver issue. But usually with a driver issue you would generally get a proper BSOD or game crash error, crash to desktop, etc.

I'll ignore that 97*C cause there is not way that sensor is accurate. But a lockup or reboot, it is possible the cpu gets too hot while gaming also.

Contact the gpu maker and see if your Radeon 250 is still under warranty.

It's sure not under warranty, also the GPU is at stock clocks and while I was gaming on a light game that sometimes froze up my pc, the temps stayed at bay. I think the issue is that one of the VRAM modules is bad, and when the GPU is under load, it hits that piece of VRAM and freaks out, causing the one of the three types of the artifacting. Sometimes on the first type it doesn't lock up my PC (i.e I can move my mouse around and stuff). The others lock up the pc and force me to shut down the pc with the power button. I was also doing AFK grinding in a game, and I got total the currency for more than when the pc started the issue, which means that the PC does lock up, but it still is active in sorts. Minecraft has the most probability of recreating the issue, I'll check temps with Speccy (not ideal I guess) And see if it reaches above 50c. (Update) I terrorized some villages in Minecraft and the issue didn't happen, I also checked temps and it never hit 50c. (Update 2) I tried running another game called Shop Titans, and before the GPU even reached 40c, the third type of the issue came up.

[Bad 💀 Motha]

Well a simple test can determine a vram issue. Use MSI Kombuster and load that up and select the VRAM test and pick the option with the amount your gpu has. Run that for about 15-30 mins and see how it does. If this fails, try down-clocking the gpu Memory using MSI Afterburner and re-run same tests.

[weed detector]

引用自 Bad 💀 Motha：

Well a simple test can determine a vram issue. Use MSI Kombuster and load that up and select the VRAM test and pick the option with the amount your gpu has. Run that for about 15-30 mins and see how it does. If this fails, try down-clocking the gpu Memory using MSI Afterburner and re-run same tests.

I'm kinda scared to use MSI Afterburner, also there's an old youtube video that has the second issue in it that I have (can't find it). Also what if I choose like 2gb of VRAM to check when I have a 1gb card, will the results come up wrong or will something worse happen? Also the reason I'm scared is that I'll break my GPU (I heard that FurMark sometimes broke peoples GPU'S), And since MSI Kombuster is based off of FurMark, it might break my GPU, I just saw a post saying that MSI Kombuster broke someones GTX 580. Do I really need to use Kombuster or should i use something like Unigine Haven?

[weed detector]

引用自 Bad 💀 Motha：

Well a simple test can determine a vram issue. Use MSI Kombuster and load that up and select the VRAM test and pick the option with the amount your gpu has. Run that for about 15-30 mins and see how it does. If this fails, try down-clocking the gpu Memory using MSI Afterburner and re-run same tests.

Also the reason I'm scared is because the broken card is the only card I have, and that means if I make a single step wrong and fry my card, then I won't be able to respond or anything. I also don't know what I might be doing in MSI Kombuster and I could easily make a single step wrong.
I just really don't want to fry my card since I've heard that someone fried theirs after using MSI Kombuster. And what if my issue doesn't even need a GPU swap?

[Set-115689]

Maybe you shouldn't be gaming with it?

You could try safe mode with networking and see if it artifacts in case it's a driver thing? Maybe something that artifacts normally but is light? Youtube?

[Viper]

I would say ..replace that GPU. But your system is kind of old anyway. I take it your not able to upgrade at this time.

[weed detector]

引用自 Set-115689：

Maybe you shouldn't be gaming with it?

You could try safe mode with networking and see if it artifacts in case it's a driver thing? Maybe something that artifacts normally but is light? Youtube?

Yeah but the artifacts are like a BSOD, They happen randomly while I'm playing and freeze up the entire PC requiring a force shutdown. Basically everything looks normal and at random times the PC will make random artifacts or a solid screen and freeze the whole PC. I cannot avoid it.

[weed detector]

引用自 Viper：

I would say ..replace that GPU. But your system is kind of old anyway. I take it your not able to upgrade at this time.

Yeah, considering my CPU bottlenecks anything that's better than a 2gb GDDR5 Gt 710 anyway. (probably) I already know my GPU is trying to trash all over me, I just want to find the root cause in case I get a replacement and sell this GPU used listed as faulty.

[weed detector]

引用自 Bad 💀 Motha：

Well a simple test can determine a vram issue. Use MSI Kombuster and load that up and select the VRAM test and pick the option with the amount your gpu has. Run that for about 15-30 mins and see how it does. If this fails, try down-clocking the gpu Memory using MSI Afterburner and re-run same tests.

I did unigine heaven and the issue didnt happen. I'm still scared to run MSI Kombuster since I really don't want to screw that up, I'm also known a bit for screwing stuff up so uhhh I guess. I just installed afterburner and apparently I have max clock. Also to what clock should i under clock? Core is at a clock of 925 mhz and the memory clock is at 1125. It's also default BIOS I think. (My PC is a prebuilt so don't blame me for not knowing a bit of basic things.)