You enabled XMP (Intel) or EXPO (AMD), hit save and exit, and now the board sits on Q-Code 00 with no display, no fans ramping normally, nothing. Sometimes it's a fresh build where you flipped XMP on right after first boot. Other times it's a RAM kit you just swapped in, or a BIOS update that reset your profile and you re-enabled it out of habit. Either way the board is dead in the water and the debug LED is frozen on 00.
I know how infuriating this is. You didn't touch voltages, you didn't overclock the CPU, you just clicked one toggle that's supposed to be safe. Let's get you out of it.
What Q-Code 00 Actually Means
00 is not a memory training error. That's the part most forum threads get wrong. On ASUS, MSI, Gigabyte and ASRock boards, code 00 means the CPU hasn't executed a single instruction. The board powered on, the VRM came up, but the processor never signaled "I'm alive" back to the chipset. Memory training happens after that handshake. So if you're stuck on 00, the memory controller inside the CPU is the thing that's failing to initialize — not the RAM sticks themselves.
When you enable XMP or EXPO, the board pushes the CPU's integrated memory controller (IMC) to run the DIMMs at their rated speed, voltage and timings. On a 13900K with a DDR5-7200 kit, that's a 2000 MT/s jump over JEDEC baseline plus a bump from 1.1V to 1.35–1.45V on VDDQ. If the IMC can't hold that, it hangs before it ever gets to POST. The board shows 00 because from its perspective the CPU is a brick.
Common real-world triggers I've seen:
- Four DIMMs populated instead of two. Four sticks at XMP speed is brutal on the IMC — most DDR5 boards cap around 5600–6000 with 4 DIMMs even when the kit says 6400.
- BIOS too old for your kit. A DDR5-6400 EXPO profile on a BIOS from launch week often won't train.
- Sticks in the wrong slots. A2/B2 matters. A1/B1 or all four slots filled without checking the manual is a classic.
- Cold boot after a power cut. The board retries training, fails, and parks on 00 instead of falling back.
The Fix — In Order
-
Power off and unplug. Flip the PSU switch, pull the cable, then hold the case power button for 10 seconds to drain the rails. Q-Code 00 won't clear on a soft reboot.
-
Clear CMOS properly. Jumper method, not just the button. Find the CLRTC/CLR_CMOS header, move the jumper from pins 1-2 to 2-3 for 15 seconds, then back. If your board only has a Clear CMOS button, hold it for 20 seconds with the PSU unplugged. Pull the CR2032 coin cell too if the board is stubborn — some Gigabyte boards keep NVRAM alive on standby power.
-
Boot with ONE stick in slot A2. Not A1. A2 is the second slot from the CPU on nearly every modern board and it's the one the IMC prefers for single-DIMM training. If it POSTs, you've confirmed the problem is memory-side, not a dead CPU.
-
Enter BIOS and leave XMP/EXPO OFF. Confirm the system boots to Windows at JEDEC defaults (usually 4800 or 5600 for DDR5). If it doesn't, stop — you have a hardware fault, jump to the section below.
-
Update the BIOS before re-enabling anything. Grab the latest non-beta from your board vendor's support page. AGESA 1.0.0.7c and ComboPI 1.1.x fixed a mountain of EXPO instability on AM5. Intel boards got similar IMC training fixes through 2023–2024. Flash it, then re-test.
-
Re-enable XMP/EXPO and immediately drop the speed one notch. On ASUS that's Ai Tweaker → DRAM Frequency. On MSI it's OC → DRAM Frequency. Take DDR5-6400 down to 6000. Take DDR5-7200 down to 6800. Boot and test.
-
If it still hangs, bump VDDQ and SA slightly. Don't touch VDIMM. On Intel 12th–14th gen, try SA 1.25V and VDDQ 1.30V. On AM5, raise VSOC to 1.20V (never above 1.30V — that's the voltage that was frying 7800X3D chips). These are the two rails that actually affect IMC stability.
If you're running 4 DIMMs, the real fix is almost always: drop to two sticks, or accept JEDEC speeds. You will not get a four-stick DDR5-6400 kit stable on AM5 or LGA1700 without a silicon lottery win.
If It Still Won't POST
At this point you need to rule out hardware. Try these in order:
- Reseat the CPU. Pull the cooler, release the socket, check for bent pins (Intel LGA) or thermal paste in the socket (AM5). Reinstall with even mounting pressure — uneven cooler torque can cause IMC failures that look exactly like 00.
- Try a different DIMM slot. If A2 fails, try B2. If both fail with a single known-good stick, the board's memory traces or the CPU's IMC is suspect.
- Swap the stick. Test each DIMM individually. One bad stick in a kit will hang training on 00 every single time.
- Check the EPS 8-pin (or 8+8) CPU power cables. A partially seated CPU power connector gives you 00 with fans spinning. Push them in until they click.
- Test the CPU in another board if you can. Q-Code 00 with a known-good PSU, known-good RAM and a cleared CMOS usually means the CPU or the socket is the problem. RMA time.
One last thing: if you're on a board with a BIOS Flashback button, you can reflash a known-stable BIOS from a USB stick without ever getting the CPU to POST. That's saved more 00-stuck builds than any other trick. Check your manual for the exact filename format — Asus wants CAP files renamed to a specific string, MSI wants MSI.ROM at the root of a FAT32 drive.
Get it posting at JEDEC first. Everything else is tuning. Don't fight the IMC on day one — you'll lose.