What is ECC memory error detection?
ECC (error correcting code) is a method for detecting and, if possible, subsequently correcting memory errors. Errors in these areas can severely limit the performance of an IT system. There are two types of single-bit memory errors: Fatal Errors and Recoverable Errors. Serious errors are caused by factors such as large fluctuations in voltage, temperature or stress on the memory. Recoverable errors occur when data is read differently than originally intended. This can be caused by voltage fluctuations on the motherboard, radiation, etc., which change bits in memory. Since the bits retain their programmed value in the form of an electrical charge, this type of disturbance can change the charge of the memory bit, thereby causing an error.
Difference Between Error Detection and Error Correction (ECC)
Error detection (parity bit)
Transmission errors can be controlled using an additionally stored parity bit. The saved byte is used to verify the other values.
Error correction (ECC)
ECC is a coding for error correction, which is in contrast to Parity check is able to correct a 1 bit error and detect a 2 bit error. The ECC process requires 6 bits on 32 bits and 7 bits on 64 bits. The ECC process is often used in memory components for server systems that require high data integrity.
Display, recognition and evaluation in toolstar®testLX
From version 5.96 of toolstar®testLX there is also the option during the memory tests to monitor the ECC error handling and, in addition to the current status, to display and evaluate the changes during the test. The prerequisite for this is that the mainboard, the firmware and the built-in memory blocks (mostly found in server systems) support this error correction and detection.
Display support and status
In the information about the RAM and the controllers you will find information about whether the function is supported on this system and the output of what values have occurred so far.
SMBios/DMI
==========
Fehlerkorrektur: keine Daten
Speicher-Controller(s)
====================
Name: Intel G41 Memory Controller
Größe: 11840 MB
Korrigierbare Fehler: 123
- ohne Modul-Informationen: 1
Unkorrigierbare Fehler: 42
- ohne Modul-Informationen: 0
Besch. Ort Typ Größe EDAC Mode Korr. Fehler Unkorr. Fehler
DIMM1 channel 1 slot 0 Registered-DDR4 x4 5920 SECDED 122 42
DIMM0 channel 0 slot 0 Registered-DDR4 x4 5920 SECDED 0 0
Displayed in test results
If the test has been activated, the memory test result will note before the memory test which values were current immediately before the test and after the memory tests which values are now current.
Speichergröße: 11.6 GB
Adressbereich: 0x0 - 0x323FFFFFF
Speicherfehlererkennung und Korrektur:
Insgesamt korrigierbare Fehler vor dem Test: 0
Insgesamt unkorrigierbare Fehler vor dem Test: 1
Adressierung bestanden
Schachbrettmuster bestanden
Speicherfehlererkennung und Korrektur:
Neue korrigierbare Fehler: 1
Neue unkorrigierbare Fehler: 0
Insgesamt korrigierbare Fehler nach dem Test: 1
Insgesamt unkorrigierbare Fehler nach dem Test: 1
Automatic evaluation
During the evaluation, the values before and after the memory test are automatically compared and, depending on the changes to the values and settings, in toolstar®testLX will automatically generate an error so that the program alerts you to this problem.
Speichergröße: 11.6 GB
Adressbereich: 0x0 - 0x323FFFFFF
Speicherfehlererkennung und Korrektur:
Insgesamt korrigierbare Fehler vor dem Test: 0
Insgesamt unkorrigierbare Fehler vor dem Test: 1
Adressierung bestanden
Schachbrettmuster bestanden
Speicherfehlererkennung und Korrektur:
Neue korrigierbare Fehler: 1
Neue unkorrigierbare Fehler: 0
Insgesamt korrigierbare Fehler nach dem Test: 1
Insgesamt unkorrigierbare Fehler nach dem Test: 1
Gesamtergebnis: FEHLER
