Gurm That's my main theory - it did some slight damage to graphics chip that causes it to generate image in a way that is unpleasant to my eyes.
I will probably be buying a new laptop and try to completely avoid installing Windows 10 on it if possible.

Gurm I've already told you my current working theory.

Gurm I seem to recall reading the first build of Windows 10 (1507, doing the hard upgrade push) broke some laptop screens by overwriting the firmware. I'll try and dig up a article later.

That would be very concerning. I realize it's a valid theory and I don't put it past MS, but that's extremely bad form. Updating firmware should not be something the OS can do. I kind of give it a pass on the Surface because MS wrote the firmware, but otherwise...

3 months later

Has anyone experienced the same problem?
Microsoft started rolling out this build last month, so some of you may already have it installed.
I've seen a few more laptops become unusable to me after 1809 update since then.
Still no idea what it does to turn devices bad permanently.

    zlhr The irony is that I find 1809 a vast improvement over previous builds. Still not voluntarily switching from 15xx though.

    zlhr Microsoft started rolling out this build last month, so some of you may already have it installed.
    I've seen a few more laptops become unusable to me after 1809 update since then.
    Still no idea what it does to turn devices bad permanently.

    Isn't this likely to be them including more recent GPU drivers from say Intel?

    • Gurm replied to this.

      AgentX20 probably. But drivers shouldn't be updating firmware. I mean, it's possible they can, but I've never once seen an nVidia or AMD/ATI driver update the VBIOS. Ever. You always have to download a crazy tool to do that. I have also never seen a stand-alone Intel driver that did so. Again, always a low-level command-prompt boot-time tool. So I really don't know what is happening here. I don't discount people's experiences, but I have no idea what mechanism could be at play here.

      Could be relation to the recent low level hardware security bugs that have been discovered

        Seagull I can understand WHY they would do it but not HOW. You generally have to single-thread the CPU and do it in a DOS prompt, basically. Unless the VBIOS is exposed to a programming interface, which is hugely expensive and almost exclusively found on server-class hardware. I just don't see this even being possible, although apparently something is happening.

        Microcode updates to the CPU are not permanent. What is a possibility however is the Intel ME (Management Engine) which is a second OS on the motherboard. The X220 is old enough that it's possible to rip out the ME. If you feel up to it, read up and give it a try. It's not a feature or service you will ever use.

        • zlhr replied to this.

          I doubt that the ME could affect display output that drastically but hey it's worth a shot.

          Try a docking station and a CRT monitor.

          • zlhr replied to this.

            Seagull
            That's my current theory.
            Since posting I have discovered that newly manufactured devices are also problematic regardless of their software.
            Is it possible that hardware and software fixes applied to address Spectre / Meltdown could have altered something that could cause this?

            ensete
            Tried to used CRT, does not help.

            Sunspark
            Sold my X220 a few weeks ago.

            I have just got an X230 in the mail w/W7 Pro. It seems 'OK' to me although I have heard there is seriously bad PWM on the screens unless it is on full brightness. Maybe software like Iris could fix that? Getting a mdp>hdmi converter so will try it out on my monitor.

            Also aren't the BIOS on Thinkpad's usually notoriously locked down? I thought most flashing tools have to show some sort of disclaimer to avoid responsibility for bricking the device if the flash fails?

            • Gurm replied to this.

              So I've now connected the X230 to my U2414h and am getting strain 🙁. (ditherig switched on and using 2012 driver)

              One thing I have noticed is the Thinkpad BIOS is a post W10 2015 version. The original BIOS for the X230 was released in May 2012.

              AFAIK the Intel Integrated VBIOS version is dependent on the motherboard BIOS - so a downgrade may work?

              This wasn't an expensive machine but don't really want to brick it - I am super curious to know if a downgrade would make it comfortable.

              • zlhr replied to this.

                diop Don't get your hopes up.
                I had one of the older BIOS versions on my X220 and DP and VGA output were always strainy (DP used with DP to HDMI converter cable).
                As for the BIOS reflash, I already sold my X220, but I doubt it would fix anything.

                Long shot, but isn't A-FRC a type of software dithering occurring at the monitor (vs. OS/GPU)?

                Only reason I bring this up: I had used two external monitors on my zbook forever. The laptop had 6bpc depth by default on its native screen, and that carried over to the external monitors. One day, I made the mistake of setting them to 8bpc, and the nVidia control panel wouldn't let me revert! Swore I had eye strain issues thereafter.

                I swear a lot of these issues coincide with the move to 8bpc+, and a lot of monitors are not yet 8bpc native (it's 6+aFRC). I wonder if something similar happened here - defaulting to a higher color depth that forced flickering.

                I have no idea if something native to the 8 bit color "palette" is what bothers my eyes or just the flickering that gets us there, but I suspect it has something to do with that on modern software & hardware. In any case, good luck :/

                diop Yes, which is why I discount this whole "Windows 10 flashed my VBIOS" idea. I could see it putting the video card into some temporary state, but draining the battery ought to fix the issue.

                • JTL replied to this.

                  Gurm I've messed up a Macbook Pro to where the internal display was flickering (like dithering but worse) when using some of my own custom methods to attempt to override low level variables in the driver, and it persisted after a reboot. Went away after holding down the command+option+shift+power for a few minutes. Scared the shit out of me.

                    dev