Announcement

Collapse
No announcement yet.

F4-3200C15D-32GTZKW stability issues

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

  • F4-3200C15D-32GTZKW stability issues

    Hi,

    I have some issues with my new build with G.Skill TridentZ F4-3200C15D-32GTZKW 2x16 GB memory.

    Intel i7-8700K
    ASUS Prime Z370-A w/latest BIOS
    Win 7 Pro x64

    It fails Prime95 within 5-40 minutes with rounding errors at the rated 3200 MHz CL15-15-15-35-2T 1.35 V settings.

    I've tried VCCSA/VCCIO voltages in the 1.0-1.3 V range without success. CPU is running stock, Asus MCE disabled.

    What is interesting is that each stick tested individually works, but they fail when paired. If I bump the tRCD and tRP I regain stability.

    My MBO sets these timings:

    CL - 15
    tRCD - 15
    tRP - 15
    tRAS - 35
    CMD - 2

    RAS to RAS Delay L - 7
    RAS to RAS Delay S - 6
    REF Cycle Time - 560
    Refresh Interval - 12480
    READ to PRE Time - 12
    FOUR ACT WIN Time - 34
    CKE Minimum Pulse Width - 8
    Write Latency - 13

    tRDRD_sg - 7
    tRDRD_dg - 4
    tRDWR_sg - 16
    tRDWR_dg - 16
    tWRWR_sg - 7
    tWRWR_dg - 4
    tWRRD_sg - 32
    tWRRD_dg - 25
    tRDRD_dr - 7
    tRDRD_dd - 7
    tRDWR_dr - 16
    tRDWR_dd - 16
    tWRWR_dr - 9
    tWRWR_dd - 9
    tWRRD_dr - 9
    tWRRD_dd - 9
    TWRPRE - 42
    TRDPRE - 12
    tREFIX9 - 108
    OREF_RI - 64

    If I downclock the memory to 2133 MHz, I regain stability. I've also tried 3400-15-16-16-35-2T 1.35V and it's much more stable but still fails after about two hours.

    I've tried the other memory slots, same results.

    Bad memory? Motherboard? CPU IMC?

  • #2
    3200-15-16-16-35-2T 1.35V is stable.

    Comment


    • #3
      1.35V is set Voltage or actual Voltage?

      Try slightly higher to see if it can improve stability. To prevent over Voltage, some motherboards have a slightly lower DRAM Voltage range compared to set Voltage.

      Comment


      • #4
        1.3530 set in BIOS, HWiNFO64 reads 1.36 V. Tried bumping it a bit higher. Didn't help.

        Comment


        • #5
          Have you tried 3000 or 2800 to see if it can work?

          Comment


          • #6
            Originally posted by GSKILL TECH View Post
            Have you tried 3000 or 2800 to see if it can work?
            At what settings? It works even at 3400 MHz with the right timings.

            Comment


            • #7
              Stable?

              Feel free to send them in to see if a new kit works better. Are you still within 30 days of purchase?

              Comment


              • #8
                Originally posted by GSKILL TECH View Post
                Stable?

                Feel free to send them in to see if a new kit works better. Are you still within 30 days of purchase?
                Yes, 3400 is stable for several hours I ran the tests. 3200-15-16-16-35-2T 1.35V is 24 h Prime95 stable (even if I overclock the CPU to 4.9 GHz)

                I'm from Europe. I am still within the 30 days but I have to return them to the shop I bought the kit from as RMA to outside my country would make me liable for VAT upon the return of the replacement kit as I'm not the original importer (dumb regulations but it is what it is).

                Is there a way you could instruct the shop I've purchased the kit from to issue me a replacement? They have hinted that they don't consider Prime95 to be a viable test (which I find absurd since a properly working system should run anything stabily) so I'm worried they'll run some basic benchmark that will show bo errors and refuse my return.

                Comment


                • #9
                  You need to contact the shop for their assistance.

                  Comment

                  Working...
                  X