Skip to content
Harrow, London · Mon–Fri 11:00–17:00 · No fix, no fee
We will be holiday from 10th of August to 2nd of September. If need any help whatsapp message us, will try to help over whatsapp if we can.
Build a PC
Home / Blog / New PC build not booting: common causes and fixes

New PC build not booting: common causes and fixes

New PC build not booting – motherboard debug LEDs and troubleshooting checks

If your new PC build won’t boot, take a breath — in the vast majority of cases nothing is broken. After years of building and repairing PCs, we can tell you the same handful of causes come up over and over again: a cable that looks plugged in but isn’t, RAM that never clicked home, a monitor in the wrong port, or a DDR5 system that is quietly doing exactly what it is supposed to do while showing you a black screen. The same applies if your PC stopped booting after you swapped a part — new GPU, new RAM, new CPU — because upgrades trip the same traps as fresh builds.

This guide is organised by symptom, so find the one that matches what your PC is doing and work through the checks in order. They are ranked from most to least likely, based on what we see on the bench and what builders report across Reddit, Tom’s Hardware and other communities every day.

Before touching anything inside the case: switch off the PSU at the back, unplug the power lead, and press the power button once to drain residual charge. Work on a hard surface, not carpet.

The 60-second sanity checks

Do these before anything else. Between them they solve an embarrassing share of “dead” new builds — and everyone has been caught by at least one:

  • PSU switch — the rocker switch on the back of the power supply must be on the I position, not O. Yes, really. It ships in the off position and it is the single most common “dead build” cause.
  • Monitor cable in the GPU, not the motherboard — if you have a graphics card fitted, your HDMI or DisplayPort cable goes into the card (lower down, horizontal ports), not the motherboard’s ports next to the USB. Plugging into the motherboard on a CPU without integrated graphics gives you a perfectly healthy PC and a black screen.
  • Wall socket and extension lead — test the socket with something else. Surge-protected extensions do fail.
  • Monitor input source — set the monitor to the correct HDMI/DP input manually. Some monitors don’t auto-switch and will happily say “no signal” on the wrong input.
  • Power lead fully seated — push the kettle lead firmly into the PSU. Half-seated leads are common on tight cable runs.

Symptom 1: nothing happens at all

You press the button and get nothing — no fans, no lights, no click. If the sanity checks above pass, the problem is almost always one of three things:

Front panel connectors. The power button on your case connects to tiny pins on the motherboard (usually bottom-right, labelled F_PANEL or JFP1), and getting the power switch pins wrong — or one pin off — is the classic first-build mistake. Check your motherboard manual for the exact layout. Quick test: with the front panel connector removed, briefly touch the two PWR_SW pins with a metal screwdriver tip. If the PC springs to life, your case wiring was the problem, not the hardware.

24-pin and EPS power. Reseat the big 24-pin motherboard cable and the 8-pin (sometimes 4+4 or dual 8-pin) EPS CPU power connector at the top-left of the board. The EPS cable is the most forgotten cable in PC building — the system is completely dead without it, and on some builds it’s hidden behind the cooler. Make sure you haven’t plugged a PCIe cable in there by mistake; they look similar but are keyed differently.

A short against the case. A motherboard standoff in the wrong position, a loose screw trapped behind the board, or the board touching bare metal will make the PSU protect itself and refuse to start. Check that every standoff under the board lines up with a screw hole, and shake the case gently to listen for loose screws. Forum threads are full of “dead” builds fixed by removing one rogue standoff.

Quick check: most motherboards have a small standby LED that glows whenever the PSU is on and connected. If that LED is lit, power delivery is fine and your problem is the power switch circuit. If it’s dark, focus on the PSU, cables and shorts.

Symptom 2: fans spin but no display

The most common new-build symptom of all: everything lights up, fans spin, and the monitor stays black. Work through these in order.

1. Wait — seriously. If this is a DDR5 build (any AM5 Ryzen or modern Intel system), the first boot triggers memory training, where the board calibrates itself to your RAM. The screen stays black, debug LEDs may cycle, and the system can restart itself several times. Crucial’s official guidance is that this can take up to 15 minutes on some systems. Every week, someone strips a perfectly healthy build apart because they didn’t wait. Give any new DDR5 build a genuine 10–15 minutes on first power-up before you touch anything.

2. Read the debug LEDs. Most modern boards have four small LEDs labelled CPU, DRAM, VGA and BOOT. Whichever one stays lit tells you where POST stopped — it’s the single most useful diagnostic on the board. DRAM stuck on means memory; VGA means graphics card; BOOT means it POSTed fine and can’t find a drive to boot from. Boards without LEDs may have a speaker header for beep codes or a two-digit POST code display; the manual decodes them.

3. Reseat the RAM. RAM needs far more force than feels reasonable — both end clips must snap fully closed on their own. Half-seated RAM is probably the number one cause of a lit DRAM LED. While you’re there, confirm the sticks are in the correct slots: with two sticks on a four-slot board, that’s almost always A2 and B2 (second and fourth from the CPU), but check the manual because it matters for training and stability.

4. Try one stick. Pull all RAM except one stick in slot A2 and try again. If it boots, add the second stick back. If it still fails, swap to the other stick — a single dead module in a kit is rare but real.

5. Reseat the GPU and its power cables. The card must click into the PCIe latch, and every power connector on it must be fully home — that includes all 6+2 pins on separate cable runs where possible, and on modern cards the 12VHPWR/12V-2×6 connector pushed in until it is completely flush. A slightly proud 12VHPWR is both a no-boot cause and a genuine fire risk. If the VGA LED stays lit and you have a CPU with integrated graphics, remove the GPU entirely and try booting from the motherboard’s video output to split the problem in half.

6. Clear the CMOS. Power off at the PSU, then either use the Clear CMOS button/jumper (check the manual) or pop the silver coin battery out for a couple of minutes and refit it. This resets the BIOS to safe defaults and clears any bad memory settings the board saved. It’s the standard fix when a system worked before a settings change and won’t POST after.

If testing does point to a dead stick, brace yourself before shopping — see our guide to why RAM is so expensive right now for what replacements cost and whether to buy or wait.

Symptom 3: it turns on, off, then on again

The PC powers up for a few seconds, shuts down, powers up again — maybe two or three cycles — and then either boots normally or gives up. Two very different causes:

If it eventually boots: this is normal. DDR5 boards routinely power-cycle during memory training, especially on first boot, after a BIOS update, after a CMOS clear, or after enabling XMP/EXPO. It looks alarming and it’s completely by design. If the long training delay on every boot annoys you (a very common AM5 complaint), look for Memory Context Restore in the BIOS — enabling it (often together with Power Down Enable) tells the board to reuse the last boot’s training data instead of retraining every time, and forum users regularly report it cutting boot times dramatically.

If it never boots: treat it as a power or memory fault. Endless rapid cycling with no POST usually means unstable memory settings (disable XMP/EXPO — see below), a failing or underpowered PSU, or the CPU EPS cable missing/loose. Overclocked RAM that passed for weeks then started cycling is a known DDR5 pattern too — drop the memory back to stock speed and retest before blaming hardware.

Symptom 4: won’t boot after a parts upgrade

The PC was fine, you changed a part, now it’s dead. The fix depends on what you swapped:

New CPU: check BIOS support first. This is the big one. A motherboard only supports CPUs its current BIOS knows about — drop a new-generation chip into an older board and you get fans spinning, CPU LED lit, no display, and nothing wrong with any part. Look up your exact board’s CPU support list on the manufacturer’s site; it shows the minimum BIOS version per CPU. Most boards from the last few years have BIOS Flashback: a USB port and button that let you update the BIOS with no CPU installed at all — download the BIOS to a USB stick, rename the file exactly as the manual specifies, and follow the flashback procedure. If your board lacks flashback, you’ll need the old CPU back in (or a shop with one) to update first.

New RAM: run it at stock first. Boot with XMP/EXPO off, confirm the system is stable, then enable the profile. If it won’t POST with the profile on, your combination of CPU memory controller, board and kit can’t hold that speed — a BIOS update often fixes it (memory compatibility improves constantly with new AGESA/microcode), or step the speed down one notch. Also: don’t mix kits. Three sticks from two different kits at different voltages is a classic forum no-boot thread.

New GPU: power and PSU headroom. Confirm every power connector is seated (see symptom 2), that you’re using the right cables for your PSU — never reuse modular cables from a different PSU, the pinouts differ and it can kill hardware — and that your PSU wattage actually covers the new card. A PSU that coped with the old card can brown-out on the new one’s transient spikes, which shows up as shutdowns under load or refusal to boot.

Any upgrade: clear the CMOS. After a CPU or RAM change especially, stale BIOS settings from the old hardware cause no-POST situations that a 30-second CMOS clear fixes. It should be a standard step after any platform-level change.

Worth knowing: if BitLocker or device encryption is on (it’s enabled by default on many Windows 11 machines), a BIOS update, CPU swap or motherboard change can trigger the recovery screen on next boot. Find your recovery key at aka.ms/myrecoverykey before upgrading — being locked out of your own drive after a successful upgrade is a horrible way to end the day.

Symptom 5: reaches BIOS but won’t load Windows

If you can get into the BIOS, the hardware has POSTed and you’re most of the way there. The usual suspects:

  • Drive not detected — check the BIOS storage page lists your NVMe or SATA drive. If not, reseat the M.2 drive (it should sit at an angle, then screw down flat) and check SATA power/data cables. On some boards, certain M.2 slots share lanes with SATA ports or run from the chipset vs CPU — the manual will say if a combination disables something.
  • Boot order — after adding drives, the board sometimes tries to boot the wrong one. Put your Windows drive first in the boot priority list, and make sure it’s the “Windows Boot Manager” entry.
  • New motherboard, old Windows install — Windows 10/11 usually survives a motherboard swap surprisingly well, but expect a driver install session, possible reactivation, and the BitLocker recovery prompt mentioned above. If it blue-screens on boot (INACCESSIBLE_BOOT_DEVICE), check the BIOS SATA/VMD/RAID mode matches what the old system used — flipping between RAID/VMD and AHCI is the usual cause.
  • USB stick still plugged in — an installer USB left in a port can hijack the boot order. Simple, common, forgotten.

Last resort: breadboarding

If nothing above works, take the build back to basics. Pull the motherboard out of the case (or at minimum disconnect everything non-essential) and run it on the motherboard box with only: CPU, cooler, one RAM stick, PSU, and video from integrated graphics or the GPU if there’s no iGPU. No drives, no case cables, no extras — short the power pins with a screwdriver to start it.

If the bare board POSTs, add one component back at a time until it stops — you’ve found your fault, and it’s often the case (a short) rather than a part at all. If the bare board still won’t POST with known-good power, you’re down to CPU, board or PSU, and it’s time for swap-testing with known-good spares — check the CPU socket for bent pins while you’re there, with a torch and at an angle.

Still stuck? When to get help

There’s no shame in tapping out. Some faults genuinely need spare parts to diagnose — you can’t prove a dead motherboard without a known-good CPU and PSU to test against, and most people don’t keep those in a drawer. If you’ve worked through this guide and you’re still staring at a black screen, or you’re not comfortable pulling a CPU cooler off a brand-new chip, stop before frustration causes real damage.

We diagnose and fix exactly these problems every week at TechSavant — new builds that won’t POST, upgrades gone wrong, and “it was working yesterday” mysteries. We’re based in Harrow and we can usually tell you what’s wrong within a day, with proper test hardware to isolate the faulty part instead of guesswork.

Book a diagnostic with TechSavant

This guide covers the most common causes based on our repair bench experience and community reports. Always consult your motherboard manual for connector layouts and BIOS procedures specific to your board — and if in doubt, ask before you force anything.

Leave a Reply

Your email address will not be published. Required fields are marked *