PC sends data, no video output from GPU or iGPU via DP or HDMI, yet DVI outperforms HDMI
PC sends data, no video output from GPU or iGPU via DP or HDMI, yet DVI outperforms HDMI
I'd like advice on how isolate the issue with my computer. I put it to sleep (suspend to RAM) last night and it had no video signal upon waking up. I thought it was the OS, but after restarting it'd post (1 short beep) and still have no video. I did several tests and debugging attempts outlined below, ultimately what worked using a old DVI to HDMI cable.
I'm baffled and suspect the mobo or CPU now. Any advice on what to try next would be deeply appreciated! I also have a recent history of issues with hard drives and a bad RAM stick that could be relevant or coincidences (also included them in the post). Seriously, it's like the computer has become haunted in the last few years. 😂
Computer Specs​
Component
Model
Available Video Ports
Motherboard
Gigabyte H170N-WIFI
2x HDMI, 1x DVI
CPU
Intel Core i5 6600
Watercooler
Deepcool Maelstrom 120T
GPU
NVIDIA Galax GeForce GTX 1060 6GB
1x DP, 1x HDMI, 1x DVI
RAM#1
16GB Corsair Vengeance DDR4-2400 1200.5MHz
RAM#2
8GB Corsair Vengeance DDR4-2400 1200.5MHz
PSU
Cooler Master V850
Case
Silverstone SG05
Storage
Type
Model
Notes
SSD/M.2 SATA
NVM Express 1.3 Adata Falcon 512GB
4yo, w/ issues
SSD/SATA
SanDisk SDSSDA240G 240GB
5yo+, boot, main OS
SSD/SATA
OCZ Agility4 128GB
10yo+
Portable/USB 3.0 (mobo)
WD WD40NDZW 4TB
1 month old, w/ issues
Shucked old portable drive
Seagate ST1000LM024 1TB
10yo+, w/ issues
Displays
ID
Original Active Port
Current Port (TEMP for debugging)
Available Ports
Dell 24"
mDP
HDMI
1x mDP, 1x DP, 1x HDMI
LG 29"
DP (daisy chained from Dell)
None
1x DP, 2x HDMI
​
The Incident​
- There's a speaker plugged into the mobo. Once I restarted the PC for getting no video, it beeped as usual indicating it posted but still had no video signal. The displays detected the cables connections but received no signal.
- The computer was set up to use the dedicated GPU for the initial display output in the BIOS/UEFI. iGPU was probably on Auto.
- I tried 3 different HDMI cables from GPU to both displays. Always detected but no signal.
- Got 4 short beeps trying to boot without the GPU. That's the code for
System timer failure
. I was trying to use the iGPU at this moment, with mobo HDMI to any of the displays. Still got detected but no video signal besides the post issue.
- Powering on without any RAM sticks gave the RAM issue beep code as expected. Still no video with only 1 stick + GPU and later after a BIOS reset, with iGPU enabled.
- Reseated PSU ATX_12 (CPU) cable just in case, but not the ATX yet because it's really hard to disconnect.
- Also tried all of above without peripherals and disks.
- Still no video after resetting the CMOS then replacing the battery, but the system started to post without beeping for system timer failure when without the GPU. (the replaced battery had 2.5V btw)
- Finally got a video signal when using a DVI to HDMI cable on mobo to any display (after failing once, but it could due the cable quality). It also works with the GPU.
It still won't work when using DP or HDMI ports from mobo or GPU, even when explicitly enabling the iGPU in the BIOS and removing the GPU.
No changes to BIOS settings have been done in one to two years before the reset.
I had to open the PC case recently. I shucked an HDD and connected directly to the SATA/PSU one or two weeks ago.
I installed the nvme SSD in the single M.2 slot available ~3 weeks ago.
I use the PC pretty much every day for several hours. I didn't notice any changes after shucking the HDD or installing the nvme.
The PSU is roughly a year old. I still have the previous one, it worked fine but it's old and a lower quality model, a Corsair CX430. I replaced it after starting to get issues with the storage, to eliminate a PSU-caused issue.
Possibly related issues​
Disks
I shucked the 1TB drive because it recently froze and suffered data corruption during a transfer but otherwise seems fine. It has no bad or pending sectors, aced tests after formatting, just showed suspicious drops in speed since the freeze. After shucking it's been working perfectly, the best speeds I've seen for it, no other freezes or performance issues so far.
I have a portable 4TB drive showing connectivity issues, even made a
thread
about it here. This one is detected by the system but won't mount 80% of the time now. Once mounted it works without disconnecting. It has a good smart, was fully written 3 times with different patterns using badblocks before I started to use it. Recently passed a scrubbing test showing no data or metadata corruption. I still have to thoroughly test it in other computers. Got no extra cable to test it, and won't shuck it for now because it's still under warranty. It's one month old, received in a RMA for another portable disk.
The previous 4TB drive started to accumulate pending sectors and a few bad sectors after nearly a year of use. Curiously it never reallocated the pending sectors even after a full write.
There was a 5TB portable before that 4TB. It was from Seagate and died suddenly with one month of use. It didn't have the click of death but wouldn't get recognised by the system at all.
The Adata SSD used to be my main OS and boot drive. It started to cause crashes and pick up media and integrity errors. A full rewrite forced it to reallocate many bad cells, it has 51% spare now but seems to have stabilized. I removed it from the computer and didn't use it for a couple of years, plugging it back and starting to use it as a scratch disk only some months ago. No new errors so far.
RAM
I recently replaced the 8GB RAM stick under warranty. It was 4yo, probably bad for at least a year, but I experienced no obvious issues until recently. All disks presenting issues have been used with the bad RAM, some of the data corruption could possibly be due it.
Motherboard
This motherboard was repaired by Gigabyte. A pin was bent when removing the protective socket plastic cover. 😬 It has been used without other PC parts showing any issues until roughly 5 years have passed after the repair. Then the first disk issues happened.
What next?​
I'm thinking of trying the old PSU, but it doesn't seem to be a power issue. Maybe reseat cables too, then try to remove the CPU, inspect the socket then reseat it? The CPU has not been touched since it was first installed, after the mobo was repaired. I have not removed the WC too, no reason because the temps are still good.
Are there any other tests I could be doing? I got a multimeter in case it's helpful.
Finally achieved a video signal via DVI to HDMI cable on the motherboard to any display (after one failure, possibly due to cable quality). It also functions with the GPU.
It might be that your PSU cannot supply enough power for the discrete GPU in your setup. Alternatively, the GPU could have failed.
Cooler Master V850
Is this PSU about a year old?
The PSU is approximately one year old. I still possess the earlier model, it operated well but is outdated and of lower quality, a Corsair CX430.
That green-labeled unit is regarded as a poor choice in the PSU industry.
Gigabyte H170N-WIFI
Try using only one stick of RAM and disabling the discrete GPU to test BIOS access. If possible, check what BIOS version you are currently running.
I thought the problem might start there, but even with DP or HDMI not working after removing the GPU and all disks, peripherals, and using just one RAM stick, it still doesn't work. Only DVI is functional, regardless of other connections.
This is the first time I encounter something like this.
I checked the receipt, it's eight months old.
That's why I hadn't tried it yet. I used to work with this build before, so I've kept it as a backup for testing.
Sorry, I forgot about it! It's F21.
I can now access the BIOS and OS, but only through DVI from either the iGPU or GPU card.
Already changed the battery thoroughly and reset several times afterward. It resolved the post-error during booting without the dedicated GPU, though the video signal problem persisted.
Is the DVI cable digital or analog? You can examine the connector closely and refer to a reference, such as an image, for example at Wikipedia. A specific picture: https://upload.wikimedia.org/wikipedia/c..._Types.svg. Everything digital relies on the DDC wire which transmits configuration data (EDID). Currently this is EDID2, and technically, older EDID is backward compatible, though it may not always function properly. Older VGA monitors began incorporating DVI, and most of them use analog DVI-I without the DDC wire. When they started adding the DDC wire (DVI-D only), the original EDID standard was retained, which often fails on newer GPUs. The DVI-I analog lacks such data or a wire/signal; it simply ignores it, meaning the monitor cannot report its own configuration. If you notice that the DVI-I works but nothing else, then the problem might be linked to auto configuration and it reverting to default "standard" settings. If this is digital DVI-D, then it’s unclear whether there should be a difference if the monitor isn’t an older CRT model.
If that applies, and the analog version functions, then a DDC wire failure or malfunction could be the cause.
Thank you for the picture, it's really useful!
It's a DVI-D connection.
Situation update, it's quite puzzling. I managed to start the operating system by booting into the OS via DVI (GPU), using the SSD/SATA Sandisk and Nvme Adata drives, and even posted my first message in this thread. After turning off the computer, I connected the SSD OCZ and restarted, eventually reaching the OS. Once logged in, it began to flicker with a certain issue. I would attribute the problem to the cable if it only occurred once during login. The glitch also disappears when the OS is shutting down or restarting. It kept happening after multiple restarts and power-offs. Sometimes, reinserting the DVI cable into the GPU after logging in resolves the issue until the next reboot. At the moment, I can't get the iGPU to display a signal (it worked with DVI earlier), likely because the GPU is properly seated. My goal was to test it, but I had an unusual idea initially.
While leaving the DVI plugged in, I also connected the HDMI cable to the GPU and used it for the second display. However, I couldn't detect any signal from the HDMI ports. But when the DVI was active, the HDMI worked perfectly. The glitch ceased when booting with both cables connected, and it stopped if I switched to using only DVI and experienced the glitch. Occasionally, the HDMI cable itself might not work—removing it could require multiple reseats for the signal to appear. I'm uncertain whether this is just a weak connection or something more complex.
I also ran a stress test by rendering a 3D scene with the GPU, displaying on HDMI in Monitor 2 and DVI in Monitor 1. So far, there have been no glitches or instability issues. I'm still confused about what's causing this behavior.
Things I haven't tried yet:
1) Unplug the OCZ to check if the glitch persists.
2) Remove the GPU to see if the iGPU also experiences similar issues.
3) Experiment with using DP and DVI instead of HDMI.
Yeah, I think it's likely the driver issue, but this feels more like a distraction. It doesn't address the main problems of no video during BIOS and grub setup, or the issue with attaching a second cable when artifacts appeared. The supposed "attach a second cable" solution might just be the force applied during installation that re-established loose connections once the system warmed up after troubleshooting.
Temperature fluctuations have been noticeable lately—just over 30°C changes in under 48 hours. The coldest night before the video issue occurred was unusually chilly for October, which is rare. I’m curious if the aging solder is contracting and expanding, causing reconnection once the system warmed up after a long period of adjustment.
HDMI and DP ports started working again after the PC had been running for a few hours, as if nothing unusual had happened before.
I’m also unsure about the HDMI ports on the mobo functioning properly. They worked when the card was new, but I haven’t used them in years. The DVI ports seem more robust, thanks to protective plastic casings that shield contacts from dust and humidity—this could explain why they were fine initially.
In any case, the brief system timer failure code has left me puzzled. There are reports linking it to bent CPU pins or simply resolving itself by re-seating the CPU or cooling components. I’m hesitant to remove the CPU right now because I don’t want to risk finding bent pins while trying to maintain the current balance that lets the system run without a backup. I’ve ordered some new thermal paste just in case another video issue arises, since I didn’t have it on hand to replace the cooler if needed.