768k Skylake Problem/Bug

[s1riker]

Originally Posted by pegnose

Try setting all your voltages to a fixed value?

Going to try that next as I got nothing else at this point.

[Brunnis]

Originally Posted by s1riker

Well just to add to the fun ...

After running HCI memtest for 48 hours with a clean bill of health and good temps all around, I figured, ok RAM's gotta be stable, so I stopped it.

I then proceeded to work all day on the PC yesterday without issue, left for a while, came back to play some GTA (maybe 20 mins at most), and left it again. Went to check on it this morning and it was frozen again.

So C-State 8 disabled, increased VCCIO and load line calibration, and new bios with MC 6A, made no difference.

EDIT: and for chalsall, I know this is not the scientific method, I'm trying multiple things at a time. But the problem is I have only 1 workstation and I need to use it as well for real work, if I were to try 1 single thing at a time, I could be at this for years. I did however test new MB (same model), new RAM (same model) and new PSU (different model) in isolation and am confident none of those are defective.

Did you try exchanging the OS drive? I've had extremely similar issues with a flakey SSD. It eventually went out, but it took 1.5 years until that happened. Before that, I got complete system freezes (usually during idle) and I could never find a way of reproducing the issue with any stress tests.

I don't know if you're the one I discussed this issue with on Anandtech's forum? If so, I've already suggested the above to you. ;)

[s1riker]

Originally Posted by Brunnis

Did you try exchanging the OS drive? I've had extremely similar issues with a flakey SSD. It eventually went out, but it took 1.5 years until that happened. Before that, I got complete system freezes (usually during idle) and I could never find a way of reproducing the issue with any stress tests.

I don't know if you're the one I discussed this issue with on Anandtech's forum? If so, I've already suggested the above to you. ;)

Yeah that was me. I haven't forgotten, but as you can imagine that involves a much bigger time commitment as I have to reinstall all my software (to be able to work) on the replacement SSD and then resync everything when I switch back, so I was putting it off in hopes something else was the cause, but it looks like I don't have much choice now.

[pegnose]

Congrats btw. to 2^(74,207,281)-1! :)

[pegnose]

Originally Posted by s1riker

Yeah that was me. I haven't forgotten, but as you can imagine that involves a much bigger time commitment as I have to reinstall all my software (to be able to work) on the replacement SSD and then resync everything when I switch back, so I was putting it off in hopes something else was the cause, but it looks like I don't have much choice now.

No, just use DriveImage XML or some other tool like Drive Snapshot and restore to another drive. You could even try drive-to-drive with DXML. That's a matter of <60 min.

EDIT: If you do this, your new drive initially might not boot. You can (with DXML) set a new disk ID and then use Windows startup repair to get it going again. Keep your old system as a backup/fallback.

And for debug purposes you could simply go dual-boot with a fresh system.

[kladner]

Originally Posted by pegnose

No, just use DriveImage XML or some other tool like Drive Snapshot and restore to another drive. You could even try drive-to-drive with DXML. That's a matter of <60 min.

These tips are much appreciated! I was unaware of those programs.

[s1riker]

Originally Posted by pegnose

No, just use DriveImage XML or some other tool like Drive Snapshot and restore to another drive. You could even try drive-to-drive with DXML. That's a matter of <60 min.

Yeah good point. I was thinking I was going to use a smaller spare SSD I had lying around, which didn't have enough space to do an image, but I just realized I no longer have that SSD, so I'm going to have to use a mechanical drive instead.

[Brunnis]

Originally Posted by pegnose

No, just use DriveImage XML or some other tool like Drive Snapshot and restore to another drive. You could even try drive-to-drive with DXML. That's a matter of <60 min.

EDIT: If you do this, your new drive initially might not boot. You can (with DXML) set a new disk ID and then use Windows startup repair to get it going again. Keep your old system as a backup/fallback.

Yep, I agree. I recently used Macrium Reflect Free to clone an old harddrive directly to a new SSD. Worked exactly as advertised and the system booted up without any fuss. The only thing that took some time was the actual copying of the content.

[chalsall]

Originally Posted by s1riker

EDIT: and for chalsall, I know this is not the scientific method, I'm trying multiple things at a time. But the problem is I have only 1 workstation and I need to use it as well for real work, if I were to try 1 single thing at a time, I could be at this for years. I did however test new MB (same model), new RAM (same model) and new PSU (different model) in isolation and am confident none of those are defective.

OK. And again, please trust I know how difficult this can be. Been there, done that (many times).

One suggestion... If you can, try a different MB by a different manufacturer with all the other components unchanged.

This /may/ be fundamental to the CPU and/or the chipset. But, as demonstrated with the "Skylake freezing bug", the "Big Boys" won't take you terribly seriously until and unless you can provide a reliably reproducible fault scenario with only one component identified as the likely cause.

[pegnose]

Originally Posted by Brunnis

Yep, I agree. I recently used Macrium Reflect Free to clone an old harddrive directly to a new SSD. Worked exactly as advertised and the system booted up without any fuss. The only thing that took some time was the actual copying of the content.

Yes, these dedicated migration tools work even better. Some of them come with new SSDs (based on Acronis e.g.) and work if at least one drive of a certain brand is involved (e.g. WD).

In any case, make sure the old partition is smaller than the new one (at least with the software I mentioned above). If it is not: use Windows or other tools to shrink the partition (e.g. EaseUs Partition Manager). Windows can't move the file table and other stuff. 3rd party apps do this during start-up.

EDIT: Ok, I see that Macrium Reflect can do even more, thanks for this valuable tip!

[LaurV]

Originally Posted by s1riker

Yeah that was me. I haven't forgotten, but as you can imagine that involves a much bigger time commitment as I have to reinstall all my software (to be able to work) on the replacement SSD and then resync everything when I switch back, so I was putting it off in hopes something else was the cause, but it looks like I don't have much choice now.

Why the bigger time commitment? Add another SSD with the same size, configure it as RAID1, run for a while or use sync software provided by the bios or by the os, take the old one out.