• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Help and questions about my RX 580 8GB

Mirkoniii

New Member
Joined
Jan 30, 2021
Messages
5 (0.00/day)
Hello, I am new here.
I bought my GPU second hand about a year and a couple of months ago and I have noticed some interesting issues with my GPU:
1. I haven't seen it go above 80watts, increasing the power limit does nothing. I thought it was because vsync was on, but turning it off still no more than 80watts.
2. I start getting artifacts at 70C (no OC). I "fixed" it with a very aggressive fan curve which keeps it below 60C.

Some of my assumptions are:
1. The previous owner mined with the card.
2. Changed it to the wrong bios or they are still mining bios.

So my questions are:
1. Can the artifacts be caused by the GPU not getting enough power?
2. Should I flash another BIOS and if yes could someone help me with it because I don't know how?
GPU.gif
BOX STICKER.jpg
GPU STICKER.jpg

My GPU-Z, the sticker on the box and the sticker on the GPU (excuse the dust).
 
Joined
Sep 3, 2019
Messages
3,088 (1.71/day)
Location
Thessaloniki, Greece
System Name PC on since Aug 2019, 1st CPU R5 3600 + ASUS ROG RX580 8GB >> MSI Gaming X RX5700XT (Jan 2020)
Processor Ryzen 9 5900X (July 2022), 160W PPT limit, 75C temp limit, CO -9~14
Motherboard Gigabyte X570 Aorus Pro (Rev1.0), BIOS F37h, AGESA V2 1.2.0.B
Cooling Arctic Liquid Freezer II 420mm Rev7 (Jan 2024) with off center mount for Ryzen, TIM: Kryonaut
Memory 2x16GB G.Skill Trident Z Neo GTZN (July 2022) 3600MHz 1.42V CL16-16-16-16-32-48 1T, tRFC:280, B-die
Video Card(s) Sapphire Nitro+ RX 7900XTX (Dec 2023) 314~465W (378W current) PowerLimit, 1060mV, Adrenalin v24.7.1
Storage Samsung NVMe: 980Pro 1TB(OS 2022), 970Pro 512GB(2019) / SATA-III: 850Pro 1TB(2015) 860Evo 1TB(2020)
Display(s) Dell Alienware AW3423DW 34" QD-OLED curved (1800R), 3440x1440 144Hz (max 175Hz) HDR1000, VRR on
Case None... naked on desk
Audio Device(s) Astro A50 headset
Power Supply Corsair HX750i, 80+ Platinum, 93% (250~700W), modular, single/dual rail (switch)
Mouse Logitech MX Master (Gen1)
Keyboard Logitech G15 (Gen2) w/ LCDSirReal applet
Software Windows 11 Home 64bit (v23H2, OSB 22631.3737)
You have it more than a year like this? The card must underperforms significantly...

Artifacts = excessive VRAM overclock or too tight VRAM timings or VRAM damage (or all of them together).
From first look this is a minning modded card 99.9% (low core clock and high VRAM clock)

This card is supposed to work at 1366MHz(Core) and 2000MHz(VRAM). The current BIOS must be heavily modded for clock speeds, low voltages and most probably tighter VRAM timings.
Is this the card (picture)?

You can try this:
Downclock your VRAM to 2000MHz or even 1950MHz and try to play games to see if the artifacting goes away. If yes its a good sign for taking further action to restore the card to default specs.
 

Mirkoniii

New Member
Joined
Jan 30, 2021
Messages
5 (0.00/day)
You have it more than a year like this? The card must underperforms significantly...

Artifacts = excessive VRAM overclock or too tight VRAM timings or VRAM damage (or all of them together).
From first look this is a minning modded card 99.9% (low core clock and high VRAM clock)

This card is supposed to work at 1366MHz(Core) and 2000MHz(VRAM). The current BIOS must be heavily modded for clock speeds, low voltages and most probably tighter VRAM timings.
Is this the card (picture)?

You can try this:
Downclock your VRAM to 2000MHz or even 1950MHz and try to play games to see if the artifacting goes away. If yes its a good sign for taking further action to restore the card to default specs.
Yes its like that for more than a year, but on userbenchmark it just says it performs under expectations not ''way below'' (which I tought it was because I didn't win the silicone lottery).
So should i remove the aggressive fan curve so it can hit 70C then downclock the VRAM to see if the artifacting (they are glitching textures) is gone?
Thanks btw, i'll try it then post again.
 
Joined
Sep 3, 2019
Messages
3,088 (1.71/day)
Location
Thessaloniki, Greece
System Name PC on since Aug 2019, 1st CPU R5 3600 + ASUS ROG RX580 8GB >> MSI Gaming X RX5700XT (Jan 2020)
Processor Ryzen 9 5900X (July 2022), 160W PPT limit, 75C temp limit, CO -9~14
Motherboard Gigabyte X570 Aorus Pro (Rev1.0), BIOS F37h, AGESA V2 1.2.0.B
Cooling Arctic Liquid Freezer II 420mm Rev7 (Jan 2024) with off center mount for Ryzen, TIM: Kryonaut
Memory 2x16GB G.Skill Trident Z Neo GTZN (July 2022) 3600MHz 1.42V CL16-16-16-16-32-48 1T, tRFC:280, B-die
Video Card(s) Sapphire Nitro+ RX 7900XTX (Dec 2023) 314~465W (378W current) PowerLimit, 1060mV, Adrenalin v24.7.1
Storage Samsung NVMe: 980Pro 1TB(OS 2022), 970Pro 512GB(2019) / SATA-III: 850Pro 1TB(2015) 860Evo 1TB(2020)
Display(s) Dell Alienware AW3423DW 34" QD-OLED curved (1800R), 3440x1440 144Hz (max 175Hz) HDR1000, VRR on
Case None... naked on desk
Audio Device(s) Astro A50 headset
Power Supply Corsair HX750i, 80+ Platinum, 93% (250~700W), modular, single/dual rail (switch)
Mouse Logitech MX Master (Gen1)
Keyboard Logitech G15 (Gen2) w/ LCDSirReal applet
Software Windows 11 Home 64bit (v23H2, OSB 22631.3737)
Yes the artifacts are glitched textures/flashing triangles, and 99% of the times the culprit is VRAM. Let’s hope it’s only from increased speed and not some permanent damage.
 
Joined
Jun 11, 2020
Messages
569 (0.37/day)
Location
Florida
Processor 5800x3d
Motherboard MSI Tomahawk x570
Cooling Thermalright
Memory 32 gb 3200mhz E die
Video Card(s) 3080
Storage 2tb nvme
Display(s) 165hz 1440p
Case Fractal Define R5
Power Supply Toughpower 850 platium
Mouse HyperX Hyperfire Pulse
Keyboard EVGA Z15
Yes its like that for more than a year, but on userbenchmark it just says it performs under expectations not ''way below'' (which I tought it was because I didn't win the silicone lottery).
So should i remove the aggressive fan curve so it can hit 70C then downclock the VRAM to see if the artifacting (they are glitching textures) is gone?
Thanks btw, i'll try it then post again.

On the bright side, games did not immediately crash which means you must have good memory.

You don't necessarily have to flash a stock bios, but that would be the most reliable way to fix it just involves some work.

You could just set your clocks in AMD software also and test for stability. Probably will need to run lower than stock speed with memory though since timings have probably changed.
 
Joined
Sep 3, 2019
Messages
3,088 (1.71/day)
Location
Thessaloniki, Greece
System Name PC on since Aug 2019, 1st CPU R5 3600 + ASUS ROG RX580 8GB >> MSI Gaming X RX5700XT (Jan 2020)
Processor Ryzen 9 5900X (July 2022), 160W PPT limit, 75C temp limit, CO -9~14
Motherboard Gigabyte X570 Aorus Pro (Rev1.0), BIOS F37h, AGESA V2 1.2.0.B
Cooling Arctic Liquid Freezer II 420mm Rev7 (Jan 2024) with off center mount for Ryzen, TIM: Kryonaut
Memory 2x16GB G.Skill Trident Z Neo GTZN (July 2022) 3600MHz 1.42V CL16-16-16-16-32-48 1T, tRFC:280, B-die
Video Card(s) Sapphire Nitro+ RX 7900XTX (Dec 2023) 314~465W (378W current) PowerLimit, 1060mV, Adrenalin v24.7.1
Storage Samsung NVMe: 980Pro 1TB(OS 2022), 970Pro 512GB(2019) / SATA-III: 850Pro 1TB(2015) 860Evo 1TB(2020)
Display(s) Dell Alienware AW3423DW 34" QD-OLED curved (1800R), 3440x1440 144Hz (max 175Hz) HDR1000, VRR on
Case None... naked on desk
Audio Device(s) Astro A50 headset
Power Supply Corsair HX750i, 80+ Platinum, 93% (250~700W), modular, single/dual rail (switch)
Mouse Logitech MX Master (Gen1)
Keyboard Logitech G15 (Gen2) w/ LCDSirReal applet
Software Windows 11 Home 64bit (v23H2, OSB 22631.3737)
On the bright side, games did not immediately crash which means you must have good memory.

You don't necessarily have to flash a stock bios, but that would be the most reliable way to fix it just involves some work.

You could just set your clocks in AMD software also and test for stability. Probably will need to run lower than stock speed with memory though since timings have probably changed.
Exactly why I said to run VRAM at stock (2000MHz) or even lower (1950MHz).

If the artifacts go away then next we can examine what alterations the BIOS has and what is the best way to proceed.
It could also have modifications to power play. For sure the Core clock is modded low and VRAM clock high. I bet both voltages also modded low.
 

Mirkoniii

New Member
Joined
Jan 30, 2021
Messages
5 (0.00/day)
I tested the downclocking and when its at 2000mhz the 3 games i tested (GTA 5, AC Odyssey, Cyberpunk 2077) crashed at about 60C. I then tested 2100mhz in cyberpunk and i was really suprised i did not get artifacts and it was stable at like 73C (I wasn't lying about it getting artifacts at 70C, I don't know why it was stable when i tested it). After that I tested 1950mhz and got minor texture glitches (which fixed themselfs when I was turned around for a bit).
I tested all the clocks in cyberpunk because it heated the GPU the fastest.
P.S I did not even have the slightest of an idea my GPU has dual bios, because I thought the bios switch would be more visible.
So should I turn off my PC before i press the switch?
I also found this forum that states that the switch to left is "mining" (which is my switch position) and "gaming" bios to the right.
 

Mirkoniii

New Member
Joined
Jan 30, 2021
Messages
5 (0.00/day)
Thank you all so much! I didn't know it was as simple as fliping a switch. I now feel like an idiot for not knowing.
I tested it. Was waiting if I could see any artifacts and didn't notice that it reached a hot 80C (fixable with fan curve) and was still stable. Power is more than 80watts when under load.
Again thank you all so much! I don't know how to mark it as SOLVED.

Gaming bios.gif
 
Joined
Oct 22, 2014
Messages
13,709 (3.83/day)
Location
Sunshine Coast
System Name Lenovo ThinkCentre
Processor AMD 5650GE
Motherboard Lenovo
Memory 32 GB DDR4
Display(s) AOC 24" Freesync 1m.s. 75Hz
Mouse Lenovo
Keyboard Lenovo
Software W11 Pro 64 bit
I tested the downclocking and when its at 2000mhz the 3 games i tested (GTA 5, AC Odyssey, Cyberpunk 2077) crashed at about 60C. I then tested 2100mhz in cyberpunk and i was really suprised i did not get artifacts and it was stable at like 73C (I wasn't lying about it getting artifacts at 70C, I don't know why it was stable when i tested it). After that I tested 1950mhz and got minor texture glitches (which fixed themselfs when I was turned around for a bit).
I tested all the clocks in cyberpunk because it heated the GPU the fastest.
P.S I did not even have the slightest of an idea my GPU has dual bios, because I thought the bios switch would be more visible.
So should I turn off my PC before i press the switch?
I also found this forum that states that the switch to left is "mining" (which is my switch position) and "gaming" bios to the right.
There is no mining Bios on the card unless it was flashed there by a user.
There is stock and performance.
 
Joined
Sep 3, 2019
Messages
3,088 (1.71/day)
Location
Thessaloniki, Greece
System Name PC on since Aug 2019, 1st CPU R5 3600 + ASUS ROG RX580 8GB >> MSI Gaming X RX5700XT (Jan 2020)
Processor Ryzen 9 5900X (July 2022), 160W PPT limit, 75C temp limit, CO -9~14
Motherboard Gigabyte X570 Aorus Pro (Rev1.0), BIOS F37h, AGESA V2 1.2.0.B
Cooling Arctic Liquid Freezer II 420mm Rev7 (Jan 2024) with off center mount for Ryzen, TIM: Kryonaut
Memory 2x16GB G.Skill Trident Z Neo GTZN (July 2022) 3600MHz 1.42V CL16-16-16-16-32-48 1T, tRFC:280, B-die
Video Card(s) Sapphire Nitro+ RX 7900XTX (Dec 2023) 314~465W (378W current) PowerLimit, 1060mV, Adrenalin v24.7.1
Storage Samsung NVMe: 980Pro 1TB(OS 2022), 970Pro 512GB(2019) / SATA-III: 850Pro 1TB(2015) 860Evo 1TB(2020)
Display(s) Dell Alienware AW3423DW 34" QD-OLED curved (1800R), 3440x1440 144Hz (max 175Hz) HDR1000, VRR on
Case None... naked on desk
Audio Device(s) Astro A50 headset
Power Supply Corsair HX750i, 80+ Platinum, 93% (250~700W), modular, single/dual rail (switch)
Mouse Logitech MX Master (Gen1)
Keyboard Logitech G15 (Gen2) w/ LCDSirReal applet
Software Windows 11 Home 64bit (v23H2, OSB 22631.3737)
As @Caring1 said, no GPU comes with mining BIOS pre-installed. This was done by previous user.
You should power down the system first before flipping the switch. Although if you flip it now it will take effect on next reboot/restart. But do it when off.

You can do this:
Before flipping the switch use GPU-Z to save the current minning BIOS to a file (rom file) with the arrow under AMD logo. Power down, flip the switch and again with GPU-Z save the file under different name. You can upload both files here or open it your self with PBE and post screenshots of each file like the one below.

1612264011304.png

And know that GPU temp is not related directly to VRAM temp. If I remember Polaris GPUs do not offer VRAM junction(in chip) temp.
 
Top