Announcement

Collapse
No announcement yet.

3600mhz CL16 Trident Z RAM constantly hard freezing on Prestige Creation x570 (3950x)

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • 3600mhz CL16 Trident Z RAM constantly hard freezing on Prestige Creation x570 (3950x)

    Hey folks,

    I also posted this over on the MSI forums but have yet to find any solution. I'm hoping someone here might be able to assist me.

    Firstly, here are my specs -
    Prestige Creation X570
    Ryzen 3950x
    2080 Ti Gaming Trio X
    G.Skill F4-3600C16Q-64GTZRC Trident Z RGB 64GB 3600MHz DDR4
    Kraken Z73
    Cooler Master 1200 Platinum PSU


    I finished building my rig a few days ago but have had several hard freezes and crashes that I managed to isolate to the RAM; or at least I'm almost certain that's what it is.
    The RAM passes various memtests at stock clock speeds (2133) but both XMP profiles result in hard freezes in windows after a few minutes of doing even the most basic of tasks. Sadly the settings that the Ryzen DRAM calculator provided me neither work in safe nor fast settings, resulting in the "memory overclocking fail" at startup. The XMP timings for the RAM are 16-19-19-39, so I went into the BIOS and manually set the frequency to 3600, the FCLCK to 1800 and made the timings slightly looser at 16-20-20-40.

    These settings worked for a day or so, and survived a barrage of tests by Aida64 for over 12 hours, Realbench, Cinebench and Memtest86. Then suddenly out of nowhere, Windows hard froze again, and again after that, and I was back to square one. I went into the BIOS and selected XMP 1, then tested each stick on its own using Karhu's Ram Tester for several hours. Each stick came back without error. I then put in two DIMMs at a time and did the same thing, also without errors. Lastly I put in all four sticks, and did the same but this time for over 8 hours (7000% RAM coverage) and again, no errors reported. After that I used Realbench's benchmark 10 times, stress test 10 times, Cinebench dozens of times, Memtest86 for several hours, rendered a 1440p 100BR 30 minute video in Premiere and AIDA64 again for several hours, all with no errors. I assumed that I had finally sorted out my issue, so I tried to use my PC normally by opening Chrome and Discord to get some work done, and Windows froze almost instantly. So over 20 hours of positive testing was worth absolute *******s.

    I then went into BIOS again, set it manually to 3600, FCLCK to 1800, slightly bumped DDRAM and SoC voltage as they were still below the max rated safe voltage for these sticks, and the freezes continued to happen. So at this point I am up the old proverbial creek without a paddle. I've tried everything I can think of and whenever I feel like I've reached a stable clock, it craps the bed again, and again, and again. Desperately need some help.

    Thanks guys.
    Last edited by Cheru; 06-16-2020, 10:14 PM.

  • #2
    Check to see what the max stable value is for each module and see if results are consistent. If they are, test for the highest stable frequency with all modules installed. With XMP enabled, does 3200 or a lower DRAM frequency work without issue? Are you using the latest BIOS?

    Comment


    • #3
      I've had the same problem GSKILL Ripjaws V DDR4-3600MHz 128GB with AMD Ryzen 3950X and ASUS PRIME X570-PRO. It wasn't stable even at SPD speed 2666MHz, let alone 3600MHz. I've tried bumping the SoC voltage all the way up to 1.2V, but no dice. However, I pulled the DIMMs out to test them on a Intel system (and it worked perfect with XMP profile), and ever since I put them back in the AMD system, there are no errors anymore, even at 3600MHz with 1.1V SoC (Uptime: 4 days 23:22:53). Before this, the system would have problems within a couple of hours.
      I think I swapped the DIMMs in the process and they pair better now. This is so sensitive. Ughh!

      Comment

      Working...
      X