Computer Turns On But No Display? Complete Fix Guide

A customer called us in October genuinely convinced he’d killed his $1,200 graphics card. He’d just upgraded from a 3060 to a 4070 Super, plugged everything back in, hit the power button — fans spun, lights came on, RGB everywhere — but his monitor said “No Signal.” Two hours of YouTube troubleshooting later, he was about to box up the card and return it.

I asked him one question over the phone: “Where is your HDMI cable plugged in?” He said the back of the PC. I asked him to look more carefully — was it plugged into the motherboard’s video output (horizontal, near the USB ports at the top of the case) or into the GPU’s video output (vertical, lower down on the case)? After a minute he answered: motherboard.

His Ryzen 7800X3D doesn’t have integrated graphics. When you plug into the motherboard’s video port on a CPU without integrated graphics, you get nothing. Move the cable down to the actual GPU output — instant display. He’d done a $1,200 upgrade and almost returned a working card because of where one cable was plugged in.

That story sounds like a one-off, but I see versions of it almost every week. The “computer turns on but no display” complaint covers an enormous range of underlying problems, from a cable plugged into the wrong port to a completely dead motherboard. The trick is to work through the diagnostic from cheapest to most expensive, and stop the moment something fixes it.

No Display Diagnostic

PC powers on (fans spin, LEDs light up) but monitor says "No Signal"

▼

1

Cable + Port

Move cable from motherboard ports (top) to GPU ports (bottom). Try different cable. Check monitor input source.

Fixes ~40% of cases

2

Reseat RAM

Power off, unplug, press both clips, push each stick until clips snap. Try one stick at a time.

Fixes ~20% of cases

3

Reseat GPU + power cables

Remove card fully, reinsert until PCIe clip locks. Re-plug 8-pin / 12VHPWR connectors until they click.

Fixes ~10% of cases

4

Replace CMOS battery (CR2032)

Pop old coin cell out, drop new one in (positive up). $2 at any drugstore. Common on PCs 5+ years old.

Fixes ~8% of cases

5

PSU or Motherboard failure

Swap PSU to test. Check for burnt smell, bulging capacitors. Debug LEDs show which stage fails.

Hardware replacement needed

Check motherboard debug LEDs: CPU → DRAM → VGA → BOOT — whichever stays lit shows where boot stalls

Why Does My Computer Turn On But Show No Display?

Common causes, in order from most likely to least:

About four out of every ten tickets we close on this complaint are monitor cable issues — wrong port, dead cable, or trying to use a port the GPU/CPU doesn’t actually support. Another two are loose RAM, which is one of the classic failure modes where the system powers on but can’t complete POST. About one in ten is an unseated or partially-failed GPU. Maybe 8% is a dead CMOS battery causing the BIOS to refuse to initialize the display. The remaining is split between dead PSUs that supply enough voltage to spin fans but not enough to power the GPU under load, dead monitors, and rare motherboard issues.

The diagnostic order matters because each step takes very different amounts of time. Cable swaps take seconds. RAM reseats take five minutes. GPU swaps take fifteen minutes if you have a spare to test. Motherboard replacement is a full afternoon. You want to rule out the cheap stuff first.

How Do You Tell If the Monitor or PC Is the Problem?

Plug the monitor into a different device — a laptop, a game console, even another computer — and see if it displays something. If it works on another device, your monitor and its cable are fine. The PC is the problem.

If the monitor doesn’t work on any device, the monitor itself or its cable is at fault. Try a different cable first (cheapest test), then a different power adapter for the monitor if it has an external brick. Some monitor failures are intermittent — the screen works for a few seconds at boot then goes dark, or the backlight is dead but the panel is still receiving signal (you can sometimes see a very faint image with a flashlight if this is happening).

Before opening the case, try pressing Win+Ctrl+Shift+B — this keyboard shortcut resets the Windows graphics driver and forces a display reinitialize. If your PC actually booted into Windows but the display output froze (common after sleep/wake), this brings it back without a restart. You’ll hear a beep and the screen will flash black briefly, then your desktop reappears.

A useful intermediate test: shine a flashlight at the dark monitor screen from a sharp angle. If you can faintly see your desktop or BIOS screen through the dark surface, the panel is receiving signal but the backlight is dead. That’s a monitor-side problem (replacement or backlight repair), not a PC problem.

Are You Plugged Into the Right Video Port?

This is the single most missed check on desktop PCs with dedicated GPUs. Here’s the deal: when you have a dedicated graphics card (NVIDIA, AMD), all your display output should go through the GPU’s ports, not the motherboard’s ports. The motherboard ports only work if your CPU has integrated graphics enabled.

How to tell which is which on the back of your desktop:

• GPU ports are lower on the case, vertically oriented (because the card sits horizontally in the PCIe slot). They look like a group of 3-4 video outputs on a metal bracket that’s separate from the rest of the motherboard’s I/O.
• Motherboard ports are at the top, horizontally oriented, integrated into the I/O shield with the USB and ethernet ports.

If you have an Intel Core CPU (any generation that doesn’t end in “F”), the integrated graphics work. If you have a Ryzen CPU, only the “G” models have integrated graphics — most Ryzen 5/7/9 desktop CPUs don’t. AMD’s high-end CPUs like the 7800X3D, 9800X3D have no iGPU per AMD’s own spec sheets. If you plug into the motherboard on those, you get zero display until you switch to the GPU port.

"On desktop PCs with a dedicated graphics card, always plug your monitor into the GPU ports, not the motherboard ports. I see this mistake roughly once a week and the fix is moving one cable. People sometimes spend hours on driver troubleshooting or order new graphics cards before realizing their CPU doesn't have integrated graphics and the motherboard port can never produce display output for them."

Are Your RAM Sticks Seated Properly?

If cables aren’t the problem, the next most common cause is loose RAM. This is one of the most reliable causes of “powers on but no POST, no display.” When even one stick of RAM isn’t fully seated, the system can’t pass its memory check and never reaches the display initialization phase.

Power down, unplug the PC, press the power button for 10 seconds to drain residual charge, then open the side panel. Each RAM stick has clips on the side that flip out when the stick isn’t fully seated. Push down firmly on each stick with both thumbs at once — you should hear or feel both clips snap closed. Don’t be gentle; RAM modules are tougher than they look and need real pressure to seat fully.

If you have two sticks, try booting with just one stick at a time. Pull both out, put just stick A in slot A2 (second from the CPU), try to boot. If no display, swap to stick B. If one stick works and the other doesn’t, you’ve identified a bad stick. If both work individually but neither works with both installed together, you might have a bad slot — try the other two slots (A1 and B1) instead.

Reseating RAM resolves about one in five “no display after powering on” tickets. It’s free, takes five minutes, and should always be attempted before assuming the GPU or motherboard is the problem.

Is Your GPU Properly Seated and Powered?

After cables and RAM, check the graphics card. Two common failure modes:

The card has come slightly out of the PCIe slot. This is more common than you’d think — the screws that hold the card’s bracket to the case sometimes loosen over time, or the card flexes downward over months of being supported only at one end. Pop the case open, remove the screw holding the GPU bracket, push the PCIe slot release tab to free the card, lift it out completely, then reinsert firmly until the release tab clicks into the locked position.

The card’s power cables aren’t fully seated. Modern GPUs have one or two 8-pin power connectors (or the new 16-pin 12VHPWR on RTX 40/50 series). These connectors are notorious for going partly in and looking fine while actually not making electrical contact. Yank each one out and push it back in until you hear a click. The 12VHPWR connector specifically has been the subject of melted-connector horror stories — make sure it’s fully seated and not bent.

If your motherboard has debug LEDs (CPU/DRAM/VGA/BOOT), watch them during boot. A solid VGA light that stays on means the system can’t initialize the graphics card. If reseating doesn’t fix that, you’re likely looking at a dead GPU or — much less commonly — a dead PCIe slot.

What Do the Beep Codes Mean?

If your case has a small speaker plugged into the motherboard’s “SPEAKER” header — a tiny black cylinder, not the audio speakers — it beeps during POST to tell you exactly what failed. Many modern builds skip this speaker, which is unfortunate because it’s the cheapest diagnostic tool you can have. They cost about $3 on Amazon if you want to add one.

AMI BIOS (most ASUS, MSI, ASRock boards manufactured after 2015):

• 1 short beep = POST passed normally. If you hear this but still have no display, the problem is the monitor, cable, or video output port — not the PC internals.
• 1 long + 2 short = video adapter failure. The GPU isn’t being detected or has failed. Reseat it; if that doesn’t help, try integrated graphics.
• 1 long + 3 short = video memory error. The GPU’s own VRAM is malfunctioning — card is likely dead.
• 3 long beeps = memory (RAM) failure. Reseat RAM, try one stick at a time.
• Continuous short beeps = power supply problem. The PSU can’t deliver stable voltage.

Award/Phoenix BIOS (some Gigabyte boards, older systems):

• 1 long + 2 short = display adapter failure.
• Continuous beeping = RAM or graphics card not seated.
• Repeating high-low pattern = CPU overheating (unusual at first boot — check that the cooler is mounted correctly).

No beeps at all with fans spinning is the most concerning pattern. It means the CPU isn’t executing BIOS code at all. On AMD systems, check for bent socket pins under the CPU. On Intel LGA systems, inspect the socket for debris or bent contacts. Could also be a dead VRM section on the motherboard — look for burn marks near the top-left of the board around the CPU power delivery area.

Modern boards with debug LEDs are more helpful than beep codes. The four LEDs (labeled CPU, DRAM, VGA, BOOT) near the RAM slots cycle through each stage during POST. Whichever LED stays lit solid after 15-30 seconds is your culprit. Some higher-end boards have a two-digit hex display — check your motherboard manual’s Q-code table to translate the number to a specific failure stage.

Could a Dead CMOS Battery Be the Cause?

This one surprises people. The small coin-cell battery on your motherboard — the CR2032, looks like a watch battery — keeps your BIOS settings stored when the PC is unplugged. When it dies, different motherboards react differently. Some just reset the clock every time you unplug. Others refuse to POST at all, including no display output, no error, no LEDs.

I worked on an ASUS Z690 Prime last October that flat-out wouldn’t POST. No error LEDs, no beep codes, no display. Customer was about to RMA it. We popped a fresh CR2032 in, thirty seconds later it booted normally. Two bucks at any drugstore.

To replace it: power off, unplug, open the case, find the silver coin cell battery on the motherboard (usually toward the middle, sometimes near the PCIe slots). Pop it out with your fingernail or a plastic spudger. Drop a new one in same orientation (positive side up — the side with the writing). Close the case, boot. If your BIOS needs to be reconfigured (date, boot order, XMP) after the swap, that’s normal — set it up once and it’ll stay until the next battery replacement in ~5-7 years.

Did You Recently Enable XMP or Overclock?

XMP (Intel) and EXPO (AMD) memory profiles push your RAM beyond its default speed. When the overclock is stable, you get free performance. When it isn’t, the system refuses to POST — no display, no error, fans spin, nothing else happens. Same thing with aggressive CPU overclocks.

The symptom is distinctive: the PC was working fine, you went into BIOS to enable XMP/EXPO or tweak clock speeds, saved and rebooted, and now the screen is black. Some boards try to boot three or four times in a row (you’ll see fans start, spin for a few seconds, stop, then start again) before giving up.

How to recover:

Clear CMOS to reset all BIOS settings back to default. Either pull the CR2032 battery for 30 seconds (same procedure as above), or use the CLR_CMOS jumper on your board, or press the BIOS reset button on the rear I/O if your board has one (common on ASUS and MSI mid-range and above).

If CMOS clear doesn’t work, try the power-cycle recovery trick: power on, wait 10 seconds for it to fail, hold the power button to force off. Repeat this three to five times. Many motherboards detect consecutive failed boots and automatically enter BIOS with safe defaults on the next attempt — which gets your display back so you can dial back the overclock.

Once you’re back in BIOS, don’t jump straight to the same XMP profile. Start one speed tier lower than rated (5600 instead of 6000, for example), and manually add 0.02V to the DRAM voltage beyond what the profile specifies. If the system POSTs stable at the lower speed, bump up incrementally. Ryzen platforms in particular are picky about memory topology — two sticks in A2/B2 is almost always more stable than four sticks.

When Is It a Dead Power Supply?

A PSU that’s failing under load can supply enough current to spin fans and light up the motherboard, but not enough to power the GPU under boot conditions. This produces exactly the “powers on but no display” symptom — fans spin, RGB lights up, but the system can’t actually initialize the graphics card.

Signs to suspect the PSU: the PC was working fine, then one day stopped showing display after no other changes. The PSU is 5+ years old. Sometimes you can hear a high-pitched whine from the PSU during the boot attempt. Sometimes the fans spin for a few seconds, slow down, then stop.

The cleanest way to test is to swap in a different PSU temporarily. If you don’t have a spare, you can do the paper-clip test we covered in our computer won’t turn on guide to verify the PSU can at least produce output at no load, though a passing paper-clip test doesn’t guarantee the PSU works under actual boot load.

If you suspect the PSU and it’s been more than 5 years, replacement is the right call regardless. A decent 650W unit from Corsair, EVGA, or Seasonic runs $70-100 with a 7-10 year warranty. Check the PSU tier list on Cultists Network before buying — not all PSUs are built equal and a cheap one can damage other components.

What If It’s Actually a Dead Motherboard?

After cables, RAM, GPU, CMOS battery, and PSU are ruled out, the remaining options narrow down considerably. Common motherboard-level failures: VRM components have burnt out, the chipset has failed, or the BIOS chip has corrupted firmware. Visible signs include bulging or leaking capacitors on the board, scorching near the VRM area (top-left of most boards), or a faintly burnt smell.

Motherboard replacement is the most expensive desktop repair short of a full rebuild. A decent mid-range board runs $150-250, plus you’ll need to reinstall Windows since the hardware change will trigger activation issues. If the reinstall goes sideways and you end up with a blue screen loop, that’s a separate diagnostic — but it at least confirms your display output is working again. If your CPU is 5+ years old at this point, it’s often more economical to do a full platform upgrade (new CPU, RAM, motherboard) rather than match a new motherboard to an old CPU that’s about to age out anyway.

This is the point where a remote diagnostic makes a lot of sense. We can help you distinguish “dead motherboard, replace it” from “actually it’s the GPU and the motherboard is fine” — and the difference between those two is hundreds of dollars in parts.

What If Nothing Worked?

You’ve tried every cable, every port, every monitor. Reseated RAM, reseated the GPU, checked the CMOS battery, suspected the PSU. And the screen is still black.

At this point we’d recommend a hands-on diagnostic. Some failures require actual touching of the hardware — checking voltages at specific points on the motherboard with a multimeter, swapping in known-good components, inspecting for physical damage under good light. Remote diagnosis can narrow the cause to one or two components but can’t physically swap hardware.

Our remote diagnostic for no-display issues runs $9.90 and typically takes 20-30 minutes. We walk through every check above, examine your specific hardware over video call, and tell you with high confidence which component is the problem. If it’s something you can DIY (cable swap, RAM reseat), we guide you through it during the session. If it needs a local repair shop, we tell you exactly what to ask for so you don’t get upsold on parts you don’t need.

Message us on WhatsApp — send a description of what your case shows when you power on (fans? lights? motherboard debug LEDs?), what’s plugged in where, and what you’ve already tried. We’ll come back with a diagnostic call.

If your primary display works but a second monitor is the one that won’t show signal, that’s a different diagnostic path — see our second monitor not detected on Windows 11 guide for the cable, port, and graphics driver checks specific to dual-display setups.

Last verified: May 2026 against Windows 10 22H2, Windows 11 24H2, and current-generation desktop hardware from Intel, AMD, NVIDIA, and major motherboard vendors.