My Ryzen system has never been fully stable with the Corsair Vengenance ram that I initially started with in 2021, which out of frustration led me to get higher end RipJaws in December 2023 (see below). Everything was fine for a couple of month now, excluding BSODs from my on testing with PBO2 that I could pinpoint and fix in the past. Then about a month ago I started having a weird problem, when I shut down windows it just rebooted. None of the software side of fixes helped. What did help in the end was unpluging the PSU from the wall and letting the condensators empty themselves, plugged it back in later and yay I can finally shut down again without having to hold the power button. That was 2 days ago.

Since then my RAM ain't stable anymore, getting weird crashes with often no bluescreens.

I bootet MemTest86+ from USB and got this: https://i.imgur.com/5rkThqm.png

Read up on it a little and disabled Global C-states (was enabled before), raised SoC voltage to 1.15v (was AUTO before), disabled USB XHCI legacy settings. Ran MemTest again and then got this: https://i.imgur.com/Cup04Mi.jpg

I then opened case and removed and reseated both sticks, swapping the first for the second and vice versa. Reran test and get this: https://i.imgur.com/QsKhieI.jpg

I've since upped the ram voltage from 1.44 to 1.45. In windows it is reported at 1.488v but still have errors when I run memtest. All other things are stock XMP, I only hard set the values secondary and tertiary values since gigabyte shows them as auto otherwise

Zentimings: https://i.imgur.com/5wAivnv.png

  • CPU: 5900x under NH-D15
  • RAM: F4-3600C14D-32GVKA (2x16GB kit)
  • MoBo: Gigabyte Aorus Master 1.2 on Bios F37f (beta)
  • GFX: RTX 4070 FE
  • PSU: Seasonic Focus PX 650W (80+ Platinum)
  • Storage: Samsung 990 PRO 2TB, 970 PRO 512GB, 970 EVO 2TB, Toshiba 20TB, 2x older WD Reds.

Can you please point me to what I am doing wrong? I thing I am doing something extremely wrong or have some part of my setup borked but can't find it. Any advice to get out of this instability hell?

1 Comment

  1. It’s either a weak IMC unable to handle tRCD 14 stable, or a weak kit not able to handle tRCD 14 without more voltage

Write A Comment