From an anonymous poster:
The point should be to encourage people to actually check their hardware. Hard drive tests are easy to do and only modestly hard to interpret properly. Memtest is hard due to lack of good tools (OEM tools vary, windows memtests doesn’t push the hardware very hard, 3rd party tools can be reliable but are sometimes buggy and may require UEFI — plus the UEFI itself has to be non-buggy). Establishing a baseline on your hardware as soon as possible is a good idea, finding defects on day 1 of the warranty is a good time to find them. You don’t want to be saying “is this a bug in memtest, or a bug with my hardware” while you are trying to rule out faulty ram.
I’ve seen people say they’ve tested hard drive and ram and still can’t figure out what’s wrong with their machine.
Running a single threaded memtest for only 30 minutes isn’t a very thorough test. (on the other hand, running a good test might be a challenge — or even impossible on some hardware)
Checked the hard drive how? “Ran chkdsk” So you checked the logical integrity of your filesystem, OK, what were the results? “There were results? …ok… After that I further stressed the hard drive (which I still haven’t properly tested) with a OS corruption test. Let me guess I was supposed to look at the results of that too?”1. Check for faulty hardware
2. Check for (and fix) filesystem corruption (not on faulty hardware!).
3. Check for (and fix) OS corruption (not on faulty hardware!).
4. Didn’t succeed? Prepare for an OS reinstall/backup-restore. Recover your files and data, make new backup (just in case). Mitigate faulty hardware (usually by replacement). Restore to alternate hard drive (don’t plug in old and alt at the same time!), find uncorrupt backup.Bad ram ranges from (in no particular order):
So bad the computer can’t boot very far (before the HD is mounted for writing) — hooray, your filesystem / OS wasn’t corrupted (before you knew something was wrong).
So subtle that you will have a trivial number a bit flips in memory during your computer’s lifetime, you will neither notice (probably) or be able to detect the fault — hooray??
Just barely enough to cause minor corruption to your filesystem / OS over several months, possible causing corruption of your backups and the filesystem(s) on your backup drives, sometimes even partially recoverable — the worst
Moderate problems, you know something is wrong and you know right away, but quick enough to avoid breaking your OS? Maybe. Hopefully windows updates don’t start installing while you are troubleshooting (or are already half installed and waiting to reboot).Hard drives are easier:
If you have evidence or good reason to suspect your hard drive has ever lost any of your data (smallest unit of measure being a sector), then you should replace the drive. Unless you don’t care about the data on the drive and the effort it would take to get back up and running should the drive vanish. Also keep in mind that for defective secondary drives IO stalls from the faulty secondary drive can cause wonky behavior for the running OS (which is on the main drive).
) Job done. Now here’s the thing sat here is a 2001 old XP machine that the Airlines cheerfully destroyed the screen on yet the HDD is unscathed after all these years although its not meteoric in performance it still works well. Lastly, as I can hear the ZZZZ’s from here should you feel the need to delve under the Hood albeit Desktop or a branded Laptop. Have a look on “YouTube” for howto’s so at least its not going to be a “shot in the dark” when you crack the case, Its handy for stuff like that and not only for the “Tide Pod” consuming fraternity as most folk think it is lol 
