Do you mind elaborating? I assume that Inspectre showed 2 radio buttons to Disable Meltdown/Spectre Protection. Because they both read Yes! in green in the Status Screen... But when you disabled them and rebooted and start Inspectre again they now report NO! in red. Did the CPUID change? Is the above right?
I have a similar older system (Z77, 4770k), also with GTX970 and I *just* updated to1809 and the cumulative update, so this is relevant. It would be helpful if we had a common test to compare, like Heaven (my favourite all purpose GPU test) or 3dMark, and not some Game. The problem is that "poor performance, "stutter", "low FPS" etc. can be subjective. We need NUMBERS and graphs, not vague statements like "my games stutter". I'd recommend MSI afterburner w/ statistics server to display actual frametimes, and then some test that everyone can easily do to see whether they also have the problem, and when and where it happens. This would be helpful.
Some continuity would help ie, Said game two or more people affected are having problem in (Diablo 3 etc) find a benchmark that uses the same Engine/api and start testing with said benchmark... Have other users on the forum / thread add results of said benchmark's runs to this thread for comparison - until you establish a baseline of some sort troubleshooting is going to be a whack-a-mole event. I kind of briefly glanced over a few of your profiles here many of you seem to be running Intel rigs (those few / 2 or 3) reporting the issues - maybe someone with AMD can verify if having this issue? - I'm personally running 1809 haven't noticed a change myself but I haven't been gaming heavily on 1809 yet so I'm probably not the best example (I'm currently stress-testing my rig atm) - KS
I understand. I ran Heaven 4.0 benchmark using the extreme preset (defaulted to 1600x900, 8xAA, DX11, windowed), and got the following: FPS: 60,9 Score: 1535 MinFPS: 25,5 MaxFPS: 121,1 I will still stress though that basically only Doom (2016) and Wolfenstein 2 exhibit the performance anomaly I mentioned in my op. This benchmark (Heaven) seemed to run about as well as I'd think it should on my system.
I unfortunately don't know how to bench Doom (2016), as it does not have a built-in benchmark - but I can still describe how to replicate the typical run I'm using to check if this "1809 performance penalty" is still present or not: - 416.34 drivers, all default settings, installed after ddu-in the old ones - fresh boot, start DOOM - in-game settings: Vulkan, Ultra quality preset, 1920:1080, 100 FOV, fullscreen, vsync disabled, best AA option (8x), show performance statistics (nightmare) - mission select --> Titan's Realm - just after loading, during black screen with mission title, my fps stays at 53-57 fps (on 1803 it was 150+) - when loaded, fps hovers at around 100 fps (1803: 140+) - not touching mouse, just moving forward, jumping down the skull-cave, after landing, about 80 fps (1803: 120+) - selecting gauss, shooting cacodemons, fps drops to 31-35 fps, stays that way for 4-6 seconds (1803: no dips below 85+) If an even more detailed description is needed, feel free to ask.
You just gave me an idea far as testing "Wolfenstein 2: The New Colossus" is built on id Tech 6 just like Doom 2016 - also has Vulkan support & has a built-in benchmarking tool. EDIT: Thanks for the info - I'll see if I have a spare GPU lying around that matches yours (970) or if not one that is known to consistently test close enough... since my 1080ti is no where near that.
Correction: I meant FRAME TIMES (in Afterburner). Because really only frametimes can give an objective idea about performance and stutter.
I've done a run in DOOM of about 3 minutes (as described 2 posts up) and logged it with Afterburner. Here's the log file. Hope this will be helpful.
Man, there are threads all over the internet talking about this issue. In general it seem to effect older chipsets/CPUs. AS to which one's I don't know. The whole focus/problem is on C:\windows\system32\mcupdate_GenuineIntel.dll which win10 1809 alters which causes the CPU problems. This is what the microcode update is effecting. https://forums.guru3d.com/threads/w...ated-with-sprectre-or-a-new-microcode.423421/
Bumping for the edit in my OP. I solved my problem and, in the end, it was not caused by 1809. It was a freak accident in which a faulty PCIe connection started acting up EXACTLY the day I did the 1809 upgrade. Man...
Glad you got it fixed. However, I hope you rubbed it dry a few times with a clean cloth or towel because WD40 leaves an oil residue behind on metal