Announcement

Collapse
No announcement yet.

Setting up G.Skill F4-3200C14D-16GTZRX

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

  • #16
    dusank,

    If the results are random, then there's not much we can do to troubleshoot until we see a pattern somewhere.

    Whiskey-Foxtrot's feedback and suggestions might be worth a try.

    Comment


    • #17
      Originally posted by Whiskey-Foxtrot View Post
      I have this same board and also the same memory kit which ran fine on the base C6H (not the extreme or wifi). When the Extreme was released it also ran fine up to BIOS 1701. Their 3008 BIOS is absolutely horrible and I get all sorts of inconsistencies with this (and other brands of memory). Memory that ran stable before, now can't pass basic tests in Windows.

      I'd give Asus some more time - for now I'm sticking with 1701 - at least I can boot up at 3200MHz and things are stable.

      If you want to stick with 3008, do not use the Stilt settings - instead go with the default memory frequency (3200MHz) and let the BIOS do the rest. I also OC the CPU, but via p-states - using the other auto OC settings never worked for me and I'd rather have p-states anyway.

      Good luck!
      I cant boot up at 3200MHz. I've got my settings from 1usmus DRAM Calculator and improved them with some from Stilts. At 3066MHz they work fine.

      I've tried BIOS 1701 but it was no improvement over 3008 for me. Read some user reports that 3008 actually improved situation for them.

      You can take a look at my settings here: Setting-up-G-Skill-F4-3200C14D-16GTZRX

      Comment


      • #18
        AIDA64 System Stability Test runs successfully for 36 hours already at 3066MHz.

        Seems that when it once boots up fine it stays stable indefinitely.

        Going to try the new beta 3101 posted by Elmor, with new AGESA 1.1.0.0. Hope it helps.

        Comment


        • #19
          With beta 3101 BIOS:

          3200MHz did not boot
          3066MHz did not boot
          2933MHz seems to run stable

          Comment


          • #20
            Got lucky, 3066MHz booted and runs stable. Until next restart ...

            Comment


            • #21
              Forgot to add to my previous post: I have 8 C6H based systems (all 3 variants - Wifi, standard, and Extreme). I do a lot of testing/reviews etc and I've always used Asus for my projects - I'm also a huge Crosshair fan, having owned (and still own) all but the 2nd Crosshair motherboard. Of course I have other vendors motherboards as well, but because I can see the consistencies between BIOS quality and regression in various BIOS updates (across all brands/models), I've come to the conclusion that Asus at this point is by far one of the most unpredictable as far as quality is concerned. To me this is alarming as between various updates, the same memory works or stops working as it previously! This is not something I am experiencing with for instance my Gigabyte or and to some extend, my ASRock motherboards. I'm not bashing Asus in any way, but their current handling of AGESA and BIOS updates has simply not been up to par to put it nicely. This includes their Threadripper boards, which I also run with G.Skill memory.

              Now, to shift to my CPUs; I've had 2 out of 8 CPUs which apparently had had bad IMM controllers causing some memory instabilities; with access to multiple systems and CPUs this was easy to narrow down. Those CPUs were since RMA'd with AMD directly and no issues were experienced with those specific systems. It could be something to look into.

              Another note: I strictly use G.Skill memory, almost exclusively for both Intel and AMD based projects; I literally have every model in the DDR3 and DDR4 line you can think of with exception of non-Samsung B-die chips, the new/upcoming GTRZX line and kits in the PC33000/4133MHz range(yet). Based on your experiences there definitely seems to be an unstable BIOS (or BIOS configuration) or (read below), a processor issue.

              I've read through your ROG forum posts and there is a lot you're trying to do all at once; new/Beta BIOS', different settings etc etc. My recommendation is 1 step at a time, test it, then move on once it's stable. If all else fails, try a different CPU if possible at all.

              Comment


              • #22
                Originally posted by Whiskey-Foxtrot View Post
                Forgot to add to my previous post: I have 8 C6H based systems (all 3 variants - Wifi, standard, and Extreme). I do a lot of testing/reviews etc and I've always used Asus for my projects - I'm also a huge Crosshair fan, having owned (and still own) all but the 2nd Crosshair motherboard. Of course I have other vendors motherboards as well, but because I can see the consistencies between BIOS quality and regression in various BIOS updates (across all brands/models), I've come to the conclusion that Asus at this point is by far one of the most unpredictable as far as quality is concerned. To me this is alarming as between various updates, the same memory works or stops working as it previously! This is not something I am experiencing with for instance my Gigabyte or and to some extend, my ASRock motherboards. I'm not bashing Asus in any way, but their current handling of AGESA and BIOS updates has simply not been up to par to put it nicely. This includes their Threadripper boards, which I also run with G.Skill memory.

                Now, to shift to my CPUs; I've had 2 out of 8 CPUs which apparently had had bad IMM controllers causing some memory instabilities; with access to multiple systems and CPUs this was easy to narrow down. Those CPUs were since RMA'd with AMD directly and no issues were experienced with those specific systems. It could be something to look into.

                Another note: I strictly use G.Skill memory, almost exclusively for both Intel and AMD based projects; I literally have every model in the DDR3 and DDR4 line you can think of with exception of non-Samsung B-die chips, the new/upcoming GTRZX line and kits in the PC33000/4133MHz range(yet). Based on your experiences there definitely seems to be an unstable BIOS (or BIOS configuration) or (read below), a processor issue.

                I've read through your ROG forum posts and there is a lot you're trying to do all at once; new/Beta BIOS', different settings etc etc. My recommendation is 1 step at a time, test it, then move on once it's stable. If all else fails, try a different CPU if possible at all.
                Thanks for sharing your mileage.

                I am not sure though what you mean by trying to do things all at once. Could you be more specific. I think that the BIOS settings I use have been already stabilized for quite some time already, I was just trying them out on the new BIOS version.

                The approach you suggest is not applicable to my situation in full, since the same BIOS settings do not yield the same results for me. Sometimes it does not boot, sometimes its unstable and freezes/reboots and sometimes it runs Stress/Stability tests for days without issues - all with the same settings applied.

                So you believe that most likely this behaviour could be caused by a faulty CPU (IMC)?

                Comment


                • #23
                  I am using these timings:



                  adjusted by the following values according to Stilt's Safe 3200MHz preset in BIOS:

                  tRAS: 34 -> 28
                  tRC: 48 -> 50
                  tFAW: 39 -> 36
                  tWR: 24 -> 10
                  tRDRD SCL: 3 -> 2
                  tWRWR SCL: 3 -> 2
                  tRFC: 560 -> 307
                  tRTP: 12 -> 8
                  tRDWR: 7 -> Auto
                  GearDown: Enabled -> Disabled


                  Further settings:

                  ProcODT = 43,6
                  RTT NOM = Disabled
                  RTT WR = OFF
                  RTT PARK = Auto

                  CAD_BUS ClkDrv = 30
                  CAD_BUS AddrCmdDrv = 30
                  CAD_BUS CsOdtDrv = 30
                  CAD_BUS CkeDrv = 30

                  DRAM Voltage = 1,4V
                  SOC Voltage = 1,15V
                  VTT DDR Voltage = 0,75V
                  Boot DRAM Voltage = 1,4V
                  VDDP Voltage = 1V
                  CLDO_VDDP Voltage = 866

                  DRAM Current Capability = 120%
                  DRAM Power Phase Control = Extreme
                  DRAM Switching Frequency = Manual
                  Fixed DRAM Switching Frequency (kHz) = 300


                  Everything else is on Default/Auto

                  Comment


                  • #24
                    That's some pretty tight settings there. This setting would only work for certain setups, so if you've got it to work, great!

                    tRAS should probably stay at 34, and tRFC can probably also be upped to around 420-480 range at the minimum. These two might help. And the tWR seems suspiciously low, but I'm not the tuning expert here. The rest seems like more dependent upon setup.

                    And with DRAM voltage at 1.4, do note that's beyond recommended for the speed you're going at, so keep your eyes on the temperature and case cooling. Might void the warranty if you're not careful.

                    Comment


                    • #25
                      All these changes including the voltages are from the Stilts preset that is included in Crosshair VI Extreme BIOS.

                      Please note that I am not running at 3200MHz with these settings as they have been targeted for, I am only able to run at 3066MHz so that makes them less tight I suppose.

                      So you suggest that increasing tRAS and tRFC might make the system more stable and increase the probability that it boots at 3066MHz?

                      I might also decrease DRAM voltage to 1,35 but I am afraid that makes the system less bootable. The modules don't seem to be too hot IMHO. How can it void the warranty? Are Samsung B-dies not able to take 1,4V without getting damaged?
                      Last edited by dusank; 12-28-2017, 11:56 PM.

                      Comment


                      • #26
                        "So you suggest that increasing tRAS and tRFC might make the system more stable and increase the probability that it boots at 3066MHz?"
                        - Yes, these might help with the stability of the system.

                        Regarding the voltage, it's just a precaution, because the warranty would be void if there are signs of extreme overheating. But if you've got enough cooling and airflow, this shouldn't be too much of a concern.

                        Comment


                        • #27
                          I've reduced the voltage to 1,35V and did not notice much of an impact on bootability. The 1,4V AFAIR was taken from Stilts profile in the BIOS.

                          Comment


                          • #28
                            I can confirm that the CPU IMC was indeed the culprit.

                            I have RMAed the CPU yesterday and today got a new one. The motherboard boots up with 3200MHz DOCP without any problems and AIDA64 System Stability Test is running for 40minutes already without any issues.

                            I would like to suggest that you add G.Skill F4-3200C14D-16GTZRX to your ASUS Crosshair VI Extreme QVL as I can confirm that it works with the motherboard flawlessly.

                            This is my first G.Skill memory (I've been buying Corsair before) and I am very satisfied with it!

                            Thanks.

                            Comment


                            • #29
                              Hi dusank,

                              Thanks for the feedback! As rare as it may be, it helps everyone here to know that the CPU IMC may sometimes be causing memory speed issues.

                              We'll make a note for our team, but they'll probably have to run through more tests before the addition to the QVL.

                              Thanks again!

                              Comment

                              Working...
                              X