You Spent $1,800 on Parts, Pressed the Button, and Got Nothing

The fans are spinning, which is the part that's confusing you

Say it's 2:40 in the morning. The side panel is off, the desk is covered in anti-static bags and zip-tie offcuts, and the pile of parts came to about eighteen hundred dollars — eight months of saying "nah, I'm good" to things you wanted.

You pressed the button. Fans spun. The RGB did its little rainbow sweep. The monitor said No Signal, thought about it for ten seconds, and went to sleep.

No beep either.

And somewhere around the fourth time you pressed power hoping for a different outcome, your brain reached a verdict: you killed something. Static, probably, when you were holding the CPU over the socket. Or a bent pin you can't see. Something expensive, done by your own hands, not fixable.

Put that down for a minute. Not because it's impossible — it isn't — but because it sits near the bottom of a list whose top entries are all some version of a connector isn't fully seated or the RAM is in the wrong two holes.

What's actually going wrong is your testing, not your hardware

Here's how the next hour usually goes. You reseat the RAM, unplug and replug the GPU, swap the monitor cable, clear CMOS, hit power. Nothing. So you do four different things.

Forty minutes later you genuinely cannot answer the question "have you tried it with one stick in the second slot and the graphics card out?" Because maybe? At some point? You're not sure, and the not-sure has now become part of the problem.

This is what the fifty-reply forum thread never fixes. It's full of people who know a great deal about hardware throwing a dozen suggestions at you at once, and you are a frightened person at 3am with no way to keep track of what you've already ruled out. The missing piece isn't knowledge. It's procedure.

So: paper. Real paper, or a note on your phone. Every attempt gets one line — what you changed, what happened. One change per attempt. You will want to skip this step. Skipping it is how people end up RMA'ing a motherboard that works.

Four boring things, and "I checked" doesn't count

The CPU power cable. Not the big 24-pin. The 8-pin — sometimes 4+4, sometimes two of them — that goes into the top corner of the board near the CPU. It's an easy one to get wrong, and it produces exactly this symptom — fans spin, nothing on screen. It's also the hardest connector to seat — worst position on the board, usually reached after the cooler's already on, and it clicks in a way that feels convincing before it's home. Pull it and reseat it properly. If your power supply is modular, check the other end too, and check you used a cable labelled CPU or EPS rather than one labelled PCIe — on some units they'll physically go into each other's sockets, and that ends badly.

Which port the monitor is in. If you have a graphics card, the cable goes into the graphics card, not the ports on the motherboard's back panel. And if your CPU has no integrated graphics, those motherboard ports were never going to output anything at all, on any day, ever.

The memory slots. Two sticks almost never go in the two slots closest to the CPU. On most boards it's the second and fourth counting outward — but "most boards" isn't good enough at 3am, so open the manual, find the memory population table, and match it exactly. Then push harder than feels polite. Both clips down.

What's under the board. A brass standoff screwed into a hole your motherboard doesn't have will short it against the case and produce exactly this symptom. Count them against the board's actual screw holes.

Take it out of the case

I know. You just spent two hours cable-managing it.

Build it on the motherboard box instead: board on the anti-static bag on top of the cardboard, CPU, cooler, one stick of RAM in whichever single slot the manual specifies, graphics card, 24-pin, CPU 8-pin, monitor cable. No storage drives. No case fans. No front panel connectors. Nothing else plugged in anywhere.

This does two jobs at once. It removes the case as a suspect — shorts, standoffs, a front panel power button wired to the wrong pins — and it means you can reach everything with both hands instead of through a hole the size of a paperback.

Start it by briefly touching the two PWR_SW pins on the front panel header with a screwdriver tip. The manual shows you which two. That's all a power button ever does.

If it posts on the box, you've learned something enormous: none of your parts are dead, and the fault is in the case. Put it back one component at a time.

The system that looks dead and isn't

A first boot can sit dark for far longer than you expect while the board trains the memory, and a first boot after a CMOS clear can sit dark longer still, because the training starts from scratch. Four sticks or a high-capacity kit slows it down again. Give it several minutes before you conclude anything.

So run a timer. Power on, leave the room, come back at five minutes. It'll feel absurd. Do it anyway.

While you're there: if you switched on the memory's XMP or EXPO profile, turn it off for now. Rated speeds aren't guaranteed on every kit-and-board pairing, and a build that won't post at the advertised speed often posts fine at the default. Tuning is a tomorrow problem.

Read the lights before you read anything else

Most current boards have four small diagnostic LEDs by the edge — MSI calls them EZ Debug LEDs, ASUS calls them Q-LEDs, and other makers have their own name. They're labelled roughly CPU, DRAM, VGA and BOOT, and they light in that sequence as each stage passes. The one it's stuck on tells you where it stopped.

Stuck on DRAM means it never got the memory working — which is a slot, a seating, or a speed problem far more often than it is a dead stick. Stuck on BOOT means the processor, memory and graphics all passed, which is genuinely good news; you're now troubleshooting a display or storage problem, not a corpse.

And the missing beep? Your board may well not have come with a speaker at all, in which case silence tells you nothing either way. Check the accessories bag before you read anything into it.

The failure that looks exactly like a dead CPU

A new processor in a board manufactured before that processor existed. The firmware doesn't recognise what's in the socket, so it never gets far enough to put anything on screen. Every part is perfect. The board just doesn't know what it's holding.

This is common on sockets that have been around for a few generations, and it's the reason "borrow an older CPU" is such a frequent answer on forums.

If your board has a BIOS flashback feature — ASUS BIOS FlashBack, MSI's Flash BIOS Button, Gigabyte's Q-Flash Plus — you may be able to update the firmware without a working processor in the socket. Board connected to the power supply, the correct file on a FAT-formatted USB stick, renamed exactly as the manufacturer's support page tells you, in the one specific rear port they specify.

ASUS, MSI and Gigabyte each document their own version on their own support pages, and the requirements aren't identical between makers — including whether a CPU needs to be present at all. So read yours, not a video for a different model. The file naming and the port both change between boards, and getting either wrong just does nothing at all, silently, which is a miserable way to spend twenty minutes.

If there's no flashback button, that route is closed and you need a supported CPU to get in.

About the static, since it's what you're really asking

Electrostatic damage happens. It's just not where the odds are, and it isn't what "fans spin, no display" typically means.

On current Intel and AMD desktop sockets the pins are in the socket rather than on the underside of the chip, so if there's bent-pin damage it's visible. Get a phone flashlight, hold it low and to the side so the pins cast shadows, and look along the rows for one that's out of line. Look — don't poke at them. And if the retention lever went down with normal effort and the CPU dropped into place without being forced, that's a decent sign in itself.

If you want a second opinion while you work through it, TrueTalk has an AI advisor called Gaming Hardware Tom, written as a gaming hardware specialist and tech reviewer covering PC building and performance tuning. He's an AI persona rather than a human technician, and he's there at 3am when no human is. Your first conversation on TrueTalk is free. Useful for talking through what your debug LED is doing and what you've already eliminated. He can't see your bench, so what you tell him is what he has to work with — which is another argument for keeping that log.

One more thing, and then go to bed if you need to. The fear here isn't really about the money, though the money is real. It's that you tried something difficult on your own and it didn't work on the first attempt, at an hour when everything feels permanent.

Almost nothing you did tonight is permanent. Connectors go back in. Firmware updates. Parts that were fine an hour ago are still fine now.

Start at the top of the list in the morning, one line at a time.