Question Black Screen and Fans Speed to 100% - Help!

Hi all,

I recently just got an RTX 3080. After several days of tweaking on Afterburner, I've now found a good downclock that has finally put an end to all the BSoD's, crashes, and kicked-to-desktops that I had in numerous games (Warzone, Cyberpunk, BFV, GTA V, Fortnite, RDR2).

I've decided to play some Dota 2 again a few days ago and I've been getting a whole new issue since. As far as i'm concerned this issue only happens on Dota 2, and at random intervals.

My screen will suddenly go black and the fans will ramp up to 100%. I can still hear the game sounds and even continue having conversations on Discord. However, the only thing I can do from there on is to restart my PC.

I've reinstalled Dota, checked for malware, reinstalled drivers and update my BIOS and chipset.

Any ideas?
 
This is a common symptom of a heat and/or power issue. What make/model is your PSU?

Are you overclocking anything )CPU, GPU, memory)?

What are the specs for this system (make/model of all installed components)?

It is true that I still occasionally will get crash/kick-to-desktops every now and then when I play Modern Warfare. But this seemed to have stopped in Cyberpunk so far. I initially found it odd that this black-screen-fan-100% issue happened while playing Dota 2 as it's a less intensive game from what I can understand. It's true that I didn't identify which fans were the ones that ramped up to full: I will update this soon.

At first, I thought the case might be choking the components as I didn't account for the GPU length upon purchase. I've had to replace one of my front 120mm intake fans with an NF-12x15 but even then the GPU is 1mm away from touching it. I, therefore, tried to play while removing the front and side panel. This is when I discovered the crashing stopped in Modern Warfare and Cyberpunk. This is when I decided to purchase the 12x15 fan to maintain a triple fan config, go with 2 PSU cables instead of daisy-chaining, and downclocked the GPU.

Alternatively, I've updated Windows 10, reinstalled GPU drivers, deleted old GPU drivers, and reinstalled the latest driver, update my BIOS and chipset, cleared CMOS, play with 1 stick of RAM only. The black-screen-100%-fan is persisting.

To answer your questions, my PSU is the Super Flower Leadex III Gold ARGB 850w. As for overclocks, there are currently 2 components:
  • CPU is overclocked to 4.0GHz.
  • RAM is overclocked to 3200MHZ CL16.
The temperatures are not particularly high whilst I am playing games. In more GPU intensive games (as I play in 1440p) such as Modern Warfare:
  • CPU Usage ~40%
  • CPU Temp ~55C
  • GPU Usage >90%
  • GPU Temp ~65C
For more CPU-bound games like Dota 2 (even in 1440p):
  • CPU Usage ~50%
  • CPU Temp ~60C
  • GPU Usage <60%
  • GPU Temp ~55C
On Idle:
  • CPU Usage ~5%
  • CPU Temp ~34C
  • GPU Usage ~1%
  • GPU Temp ~40C
As it currently stands, my GPU is downclocked to the following in order to avoid crashes in other games:
  • Power 85%
  • Core Clock -150MHz
This is my full built list:
Ryzen 7 2700X
Noctua NH-D15S
ROG Strix B450-F Gaming
Gigabyte RTX 3080 10GB Aorus Master
G.Skill Trident Z RGB 2x8GB 3200Mhz
Adata SX8200 Pro 512GB ◾ KLEVV Neo N610 1TB
Seagate Barracuda 2TB ◾ WD Blue 2TB
NZXT AER 140mm & Hue+ ◾ Noctua NF-A12x25 PWM
Super Flower Leadex III Gold ARGB 850w
Fractal Design Meshify C Dark TG

I hope all of this information helps.

Edit: Added more info
 
Last edited:
If you revert to stock speeds (drop memory to default 2133), is the system stable?

I will try this tonight. I have had RAM issues in the past. I noticed this for the first time when I was upgrading from the stock CPU cooler to the Noctua cooler last year. Though I never had issues before, during this upgrade, I was getting a motherboard RAM debug light and my PC didn't want to post. I had to clear CMOS, run a single stick at stock 2133 speeds to get it to post and was later to add the second stick of RAM and re-OC to 3200MHz.

So I think you might be on to something. I will get back to you once I've done some gaming on 2133MHz.
 

COLGeek

Cybernaut
Moderator
I will try this tonight. I have had RAM issues in the past. I noticed this for the first time when I was upgrading from the stock CPU cooler to the Noctua cooler last year. Though I never had issues before, during this upgrade, I was getting a motherboard RAM debug light and my PC didn't want to post. I had to clear CMOS, run a single stick at stock 2133 speeds to get it to post and was later to add the second stick of RAM and re-OC to 3200MHz.

So I think you might be on to something. I will get back to you once I've done some gaming on 2133MHz.
Someone is always here to help. Good luck.
 
Hello, so I am back after some gaming. To confirm, I've set everything in my BIOS to Auto today, which defaulted my CPU speed to vary between its stock speed of 3.7GHz to 4.0GHz. The RAM is also set to Auto, which defaulted it to 2133MHz @ 1.2v. I've also reset my downclocks on MSI Afterburner so the GPU is currently running at stock speeds. It seems as though the issue is less frequent but still present as I experienced another black-screen-100%-fan on Dota 2. So far I've had no crashes/kicks or BSODs on Cyberpunk with this configuration but I've not had any long (more than an hour) sessions.

Bonus testing I've done: When I downclock or use stock clocks on the GPU. TechPowerUpGPU app shows VRel and PWR respectively as a limit. So nothing unusual from what I can tell.

Side Note: Small update also on my previous configuration (from my first initial post): the crashes came back as well on the same night I left our conversation. So I was experiencing the crash/kick-to-desktops and occasional BSODs (with no particular error code) while Warzone and Cyberpunk on top of the black-screen-100%-fan on Dota 2 only.

Edit: it is as frequent
 
Last edited:
I just wanted to provide a small update on the situation. I don't know why I didn't think of doing this sooner but I've popped my old GTX 1060 6GB back in and the system was running fine (even after re-overclocking my CPU to 4.1GHz and and my RAM to 3200MHz). This led me to believe that issue came from either the 3080 itself or the PSU being defective.

I then put the 3080 back in and I decided to play with my side and front panels open but no luck either. I've since gotten my card reimbursed. I think I will opt for one of AMD's offerings in the near future or perhaps an RTX 3080 model, which uses one of the more stable* capacitors. If the problem reproduces itself, I will get the PSU replaced as well.

If the problem still persists after all of that. I will reach out to you and either re-open this thread or post a new one.

Appreciate all of your help so far. It's helped me to narrow down the issue.