Intel® SSDs: Important SMART Attribute Indicators
| Note | Intel's NAND SSD business has been acquired by SK hynix and is now Solidigm™. See Customer Support changes for more details. |
This article is to educate customers of Intel® SSDs on how to determine the health of their SSDs using a subset of important SMART or similar attributes.
| Note | Some attributes may be product specific and not appear in the SMART readout. |
SATA SSD Example
The SMART readout includes
- Normalized values
- Raw values,
- Threshold values
- Other values
Note that in some cases, the Raw value of the attribute matters and in other attribute cases,the Normalized value of the attribute matters. Also note that only pre-fail attributes have a Threshold value, and the attribute value can go from good to bad in either direction: increasing or decreasing in number/value. In the example below, the Available Reserved Space attribute value for a good SSD is anywhere between 100 and 11 (10 and below is bad), whereas BBh is best at raw value of 0. (See table.) Some attributes such as AFh work differently. Please ignore the rest of the information in Status/Threshold/Worst.
The table below shows the most important attributes on SATA SSDs, with corresponding look-at values, threshold values, and also acceptable condition and best case conditions. If the attributes are within acceptable conditions, the SSD is in a good state. If they're outside of acceptable conditions, contact Customer Support.
| Important Attributes | Attribute Name | Look-at Value | Threshold | Acceptable Condition |
| B8h | End-to-end error detection count | Normalized | 90 | 91-100 (100=best) |
| E8h / AA (Duplicate attribute) | Available reserved space | Normalized | 10 | 11-100 (100=best) |
| E9h | Media wearout indicator | Normalized | 1 | 2-100 (100=best) |
| AFh | Power loss protection failure | Normalized | Fail=below 10 | Shows huge number |
| 05h | Reallocated sector count | Raw | 1000 | 0-999 (0=best) |
The example below shows a good SSD. The look-at values per important attribute are highlight in yellow, which are all in a best-case state.
PCIe® SSD Attributes
These attributes can be analyzed in the example below. Some of the most important ones are highlighted in yellow. The Critical Warnings indicator is actually the most important one, as it captures all other important indicators. Unsafe Shutdowns should make sense for the SSD users, or there could be an issue with the OS driver (not sending safe shutdown signals). Temperature indicators indicate if cooling is sufficient. Percentage Used indicates media wear, and a value between 1 and 99 is acceptable, whereas 100 is not. Fewer Media Errors indicates better SSD health. If any of the attributes mentioned here cross acceptable threshold limits, they result in Critical Warnings. If you see any Critical Warnings, contact Customer Support.
