Page 1 of 1

interpreting SSD surface test results

Posted: 2020.09.15. 21:12
by molnart
Hi,
i have an ssd that was giving me headaches on a different computer, i have moved it to my main PC to do some extensive testing with hdsentinel.

The disk write test gave me some slower sectors towards the end of the disk, so i have decided to do a full surface reinitialization, that took 12 hours to finish with huge slowdowns, especially towards the end of the test. see below (fyi: the test was performed from the back of the disk)
20200914-070102_RI_Crucial_CT525MX300SSD1_163813FCB384_M0CR070-surface-full.jpg
20200914-070102_RI_Crucial_CT525MX300SSD1_163813FCB384_M0CR070-surface-full.jpg (836.57 KiB) Viewed 3898 times
afterwards i did a standard write test, with the following results:
20200914-140440_W_Crucial_CT525MX300SSD1_163813FCB384_M0CR070-surface-full-st.jpg
20200914-140440_W_Crucial_CT525MX300SSD1_163813FCB384_M0CR070-surface-full-st.jpg (563.64 KiB) Viewed 3898 times
i have presumed the disk is dead, i was not even able to partition it.

today i have decided to do one last write test, that resulted like this, actually very similar to the very first hdsentinel test i performed
Result today
20200915-204511_W_Crucial_CT525MX300SSD1_163813FCB384_M0CR070-surface-full-st.jpg
20200915-204511_W_Crucial_CT525MX300SSD1_163813FCB384_M0CR070-surface-full-st.jpg (523.77 KiB) Viewed 3898 times
my question is, how do i interpret these results? is something wrong with the disk, or the slowdown towards the end (and the disastrous results of the reinitialization test) are caused by some controller/cache overflow causing slowdowns? how to test it further?

why i am concerned is, that the "headaches" mentioned above occur on the other computer with a brand new drive as well, so i am not sure the disk was the culprit, nor that this Crucial SSD is actually bad or not.

Re: interpreting SSD surface test results

Posted: 2020.09.16. 10:22
by hdsentinel
> i have presumed the disk is dead, i was not even able to partition it.

Sounds so interesting. Generally yes, while the health is not 100% according the bottom of the surface test images, the test seems complete without problems.

Yes, because of internal device cache and similar technologies, it is expected to see that on flash storage (SSDs, memory cards, pendrives) a write type test starts relatively quickly - and then, as we advance we got worse and worse write speed. This is completely normal and expected behaviour.

What kind of error message do you see when you try to partition/format the SSD? After the Reinitialisation, I'd expect no problems, generally then the SSD should be partitioned without problems.

> why i am concerned is, that the "headaches" mentioned above occur on the other computer with a
> brand new drive as well, so i am not sure the disk was the culprit, nor that this Crucial SSD is actually bad or not.

Is your other SSD made by Crucial ?
There are some compatibility issues of Crucial SSDs with some motherboards / chipsets (and their older drivers). Some other topic (eg. https://www.hdsentinel.com/forum/viewto ... 32&t=12514 ) discuss for example new problems.
So not sure, but maybe the chipset drivers would require an update as it may improve compatibility with this (or these) SSDs in general.

If possible, please use Report menu -> Send test report to developer option about this SSD (especially if you still have problems when you try to format it), would be nice to see the current situation (complete self-monitoring information and possible chipset/driver) as it may give further thoughts.

Personally I'd also try the following:

1) select Disk menu -> Surface test
2) select Write+Read test but do not start yet
3) select the Configuration tab in this window and on the right, select Write pattern = Random data

then start the test. It will overwrite all sectors with random data and then verify if the data could be read back and we receive back the REAL data what we wrote. This would confirm if all sectors of the SSD can really store and hold the information we wrote previously.
(The Reinitialise Disk Surface performs slightly similar, but that writes all zeroes in the last step and verify/confirm that we read back instead of any other data, to ensure that all sectors properly erased and clean. By the above method, you can verify if any non-empty data could be written, read back - without damages/corruption).

Re: interpreting SSD surface test results

Posted: 2020.09.17. 23:34
by molnart
now the drive seems to work properly. the error when trying to format it was something about not supported command or something, i did not do a screenshot and a restart solved the problem so i can't reproduce.

also testing just the end of the disk (that showed slower write speeds) has now normal speeds.

Re: interpreting SSD surface test results

Posted: 2020.09.21. 09:37
by hdsentinel
Thanks for the update, good to hear ;)