UEFI POST Diagnostic Codes: A Complete and Practical Guide

When a computer or server is turned on, the first thing that comes into play is not the operating system, but the firmwareThat is, the small block of code that resides on the motherboard and decides whether the computer is ready to boot. In modern computers, this role is usually played by UEFI, the successor to the classic BIOS, and one of its essential tasks is to execute the POST (Power-On Self Test), a series of checks that, if something goes wrong, generates error codes, beeps, and messages that often sound like gibberish if you don't know how to interpret them.

Los UEFI POST codes They are the key clue to knowing what is failing: from a faulty RAM module to a poorly seated graphics card, a damaged hard drive, or a communication error with the management controller (BMC, IMM, SP, etc.). Understanding what they mean, how they are logged, and what actions each manufacturer recommends (Lenovo(IBM, Dell, Oracle, etc.) is essential for both technicians and advanced users who want to diagnose their equipment without getting lost in unnecessary technicalities.

What are UEFI and POST and what role do they play when starting a computer?

UEFI UEFI (Unified Extensible Firmware Interface) is the standard that has replaced the traditional BIOS in most modern computers and servers. Unlike the classic BIOS, which operated in 16-bit mode and was tightly tied to the motherboard's ROM, UEFI can reside in any non-volatile memory and runs in 32-bit or 64-bit protected mode, allowing it to offer more features, better performance, and advanced security options.

POST The Power-On Self Test (POST) is the "medical check" that the firmware performs on the hardware as soon as the computer is turned on. During this phase, components such as the processor, RAM, chipset, graphics card, keyboard, disk drives, and other essential devices are detected and initialized. If everything is working correctly, the firmware passes control to the operating system loader; otherwise, it generates error codes, on-screen messages, and beep patterns to indicate which part has failed.

The POST process is carried out very specific testsThe CPU is checked, the firmware itself is verified, the configuration saved in the CMOS memory is read, the system timer (internal clock) and the DMA controller are initialized, the RAM and cache are checked, additional firmware modules are loaded, and basic peripherals such as the keyboard and devices are detected and configured. storage.

If the POST finds a mild anomalyTry to continue with the BootBut as soon as it detects a critical error (for example, no usable RAM, the graphics card is not responding, or the CPU fails), it stops and displays one of these signals: an on-screen message if the video subsystem is working, a sequence of beeps with a specific pattern, and/or POST codes sent to internal ports or logged in the system management controller.

When everything is correct and there are no detectable faults, many firmwares emit a single short beep which, traditionally, has been interpreted as confirmation that the POST has finished without problems and the computer is going to hand over control to the operating system.

UEFI POST diagnostic codes on servers (IMM, SP, BMC, etc.)

On many modern servers, UEFI POST diagnostic codes don't just remain as screen messages or beeps, but are They record in detail in the integrated management modules. Lenovo, for example, stores these codes in the IMM event log; other manufacturers use drivers called SP or BMC accessible through interfaces such as Oracle ILOM or similar.

When a POST event occurs, manufacturers' guides usually include a field called “User response”There you will find step-by-step instructions on how to troubleshoot the problem. The idea is to follow the instructions in the order presented, and if the error persists after completing all the steps, contact the appropriate technical support (Lenovo Support, IBM Support, etc.).

The UEFI/POST server manuals compile extensive lists of error codes accompanied by descriptions and possible causes. For example, they can indicate if the error is related to the memory subsystem, the PCIe bus, the I/O controller, thermal problems, SAS/SATA hard drives, CMOS battery failures, password errors, etc.

An important part of the diagnosis involves reviewing the SP or IMM event logThere you can find more detailed messages than those displayed on the screen, often with additional information: affected port or slot, exact type of bus error, memory module involved, thermal thresholds exceeded, and even histories of repetitive errors that help identify intermittent or soon-to-fail components.

In some cases, a POST error may require a simple management controller resetFor example, certain internal communication failures between the main firmware (BIOS/UEFI) and the SP/BMC can generate warnings such as "BMC Not Responding", and the recommended solution may be to restart the SP or, if that does not resolve the issue, escalate the case to support to rule out a physical failure on the board or in the management module itself.

Examples of common POST messages on UEFI servers

In server environments it is common to find errors associated with IOH during the last boot. Messages such as “Uncorrectable Error Detected on Last Boot: IOH(0) Protocol Error”, “IOH(0) QPI [x] Error”, “PCI-E [x] Error”, “ESI Error”, “Thermal Error”, “Miscellaneous Error” or “VT-d Error” indicate problems detected in the input/output concentrator or in the links it manages.

In all these cases, the typical recommendation is review the fault management function and the SP event log in Oracle ILOM or another equivalent console for more details. These logs typically specify whether it is a one-off failure, a thermal condition, a QPI link error, a problem with a specific PCIe port, or a conflict with I/O virtualization functions (VT-d).

Another common message is “Hard disk error” associated with SAS or SATA failures. The firmware usually indicates that an error has occurred in a disk drive and advises checking the SP log to see which bay, port, or LUN is causing the problem, as well as checking the status of the disks on the corresponding controller.

Messages like “Bad PBR sign” These errors point to problems with the disk's partition table. In these cases, the cause is usually a missing or corrupted partition table on the drive. The solution involves recreating the partition tables with appropriate tools (for example, the Oracle Solaris formatting utilities or fdisk en Linux) and reinstall the system if necessary.

Errors of the following type appear in the memory subsystem “RAM R/W test failed”These errors indicate that the RAM read/write test during POST has failed. Again, it is recommended to review the SP event log to locate the specific affected bank or module, and then proceed with a physical check of the memory modules (reseat them, test them individually, or replace them).

Notices related to the battery and CMOS configurationsuch as “CMOS Battery Low” or generic CMOS errors. These messages indicate that the battery responsible for powering the CMOS memory—where the time, date, and part of the configuration are stored—is depleted or failing, and that it is advisable to check the event in the SP and, normally, replace the battery.

In the security section, messages such as the following appear: “Password check failed” When the password verification configured in the firmware fails, these events are also logged in the SP/IMM in a managed environment to track failed attempts to access the system configuration.

Meaning of POST beeps in BIOS Award and AMI

In addition to on-screen or register codes, many BIOSes still use beep patterns to indicate errors, which is especially useful when the system fails to initialize the video. The combination of short and long beeps helps identify whether the problem lies with the RAM, the processor, the keyboard, the graphics card, or the BIOS itself.

On systems with modern Award BIOS, the beeps are usually focused on video problemsThus, a long beep followed by two short beeps indicates an error in the graphics card or one of the video devices; the recommendation is to check that the card is properly seated in its slot (PCIe, AGP in older computers) and try another one if possible. Other patterns usually point to memory errors.

At Award BIOS table It also details combinations such as one short beep (correct startup), two short beeps (CMOS problem solvable by resetting the configuration by removing the battery or moving the appropriate jumper), one long beep and one short beep (problems in the motherboard or RAM), one long beep and three short beeps (keyboard), and so on, with recommendations ranging from repositioning modules to replacing components.

The BIOS AMI They have a slightly more extensive table of numbered beeps. For example, one beep indicates a memory update error, two a parity error, three a fault in the first 64 KB of RAM, four a non-functional system timer, five a processor fault, six an A20 fault, eight read/write errors in video RAM, nine an incorrect ROM checksum, ten a problem in the CMOS close register, and eleven a cache fault.

The way they act is usually similar in most cases: reposition modulesTry replacing a different component, or if the problem points to the motherboard or BIOS, send it in for repair or have the firmware chip updated. Many manuals also recommend using soft erasers to clean the contacts of RAM modules or graphics cards when a poor connection due to dirt or oxidation is suspected.

In practice, one can distinguish two “families” of beeps They're fairly easy to recognize: continuous, short, or long and repetitive patterns usually point to RAM or power supply problems; while non-repetitive melodic sequences tend to indicate graphics card failures or other specific devices. Understanding this pattern helps pinpoint the suspected component.