Page 6 of 268 FirstFirst ... 4 5 6 7 8 16 56 106 ... LastLast
Results 51 to 60 of 2679
  1. #51
    ROG Member Array
    Join Date
    Apr 2015
    Reputation
    10
    Posts
    18

    Quote Originally Posted by Raja@ASUS View Post
    That's what the CPU is programmed to request under XFR by AMD. 1.45V is normal. If your CPU can handle less, then set less.
    That would be up to AMD, not us. Right now, the current AGESA microcode doesn't even have user control over the sub-timings, let alone more ratios.
    Typical AMD tried to burn their CPUs since Athlon XP days. yes 1.35v will get me long way.
    And fair enough on memory speeds, I guess we need to kick AMDs arse a bit, for them to wake up

    But regarding the temps, I am now running at stock frequencies(turbo is OFF, so 3.6ghz)@1.35v and I have CPU loaded 100% with some AVX code. I locked the fan to 65% top speed which results in 1000rpm. HWinfo reports CPU temps as 75C, which is completely wrong since I can touch easily base of noctua cooler and it is not even warm to the touch. From my previous experiences 75C is very hot to the touch. Contact between CPU and heatsink is perfect: https://flic.kr/p/RU987E

    Is this an effect of that reported temperature offset which AMD for some reason put in? I have no other explanation otherwise.

    Thanks for replies, by the way

  2. #52
    New ROGer Array jrbris PC Specs
    jrbris PC Specs
    MotherboardROG Crosshari VI Hero
    ProcessorAMD 1800X
    Memory (part number)Corsair vengence LPX 3200 mhz
    Graphics Card #1sapphire tri-x 390x 8 gig
    Graphics Card #2sapphire tri-x 390x 8 gig
    Sound Cardcreative Zx Series
    MonitorLG 29 inch Ultrawide
    Storage #1Samsung Evo M.2 500 gig
    Storage #2Samsung SSD 250 gig
    CPU CoolerRGB 3.0 Big water kit 390mm Radiator
    CaseWD Black 3TB
    Power SupplyCorsair 1000W
    Keyboard Raver Deathstaker
    Mouse Logitech G502
    OS Windows 10 Pro 64bit

    Join Date
    Mar 2017
    Reputation
    10
    Posts
    2

    ramblings

    Quote Originally Posted by Demoniacstar View Post
    i have noticed on my rig when SoC voltage is too high i will get errors just like that ......i got above 2666 by setting 4 ghz on all cores running 1.375 or 1.378 vcore ...i can't remember what the settings says in bios right now ...i switched up to 4.1 ghz just now .....but anyway ....i set my 4ghz setting up and set ram to 2666 and i also set ram voltage to 1.30 in bios even though my ram calls for 1.35 for 3000 mhz because llc will bring it to 1.373 in windows .....very very important piece of this puzzle is SoC voltage .....start with a flat 0.95 ....you can then raise bclk one click at a time and test....but try keeping your cpu to a known stable clock while raising bclk
    i do not know why but to much and over volting the SoC will have the same effect as not having enough Soc very hard to determine the right amount that is why i use Aida64 cache mem test so i can see the effect the voltages are having on the memory systems of this set up.....i can not post pics or i would be posting the cache mem test results showing the big ...VERY BIG difference even just one click too much SoC will cause...
    i got my ram to 2725 with 102.200 bclk and timings of 14-14-14-14-32-cr1(1001 bios).....and it seems stable as a rock on ram still working that 4.1 ghz setting trying to get it stable ...i do not like the amount of vcore i need to get it stable but raising SoC more then 0.96 on my system seems to have harsh effects on overall scores in the cache mem test...thats why i say this beast of a cpu has a very strong imc and it does not need much SoC voltage at all for the speeds i am currently running ......with this set up i get 400+GB/s across the board on my l3 cache in aida64 if you're under that at 4/4.1 ghz you might be pumping to much SoC voltage.....SoC also seems to affect l2 cache as well but not as much as l3 it seems ......i have also noticed the 1.8 pll voltage if to high can cause crashing ....
    i do not know why but this voltage if left on auto seems to want to run 1.87 volts....i turned it down to 1.85...i tried 1.8volts and it did not like that ...lol.
    i do not know if my ramblings are even helpful..lol....but i love the hell outa this system and i want to help all who have one enjoy it as much as i do!...

    hopefully this new microcode with a bios update really opens this beast up!!!! not that it really needs it ...lol.....even at the settings im running now this thing is very very impressive to me!..... but yeah man ...3000/3200 or higher ram will make this thing untouchable by intel man .....and i cant wait to see the faces on my intel fan boy friends when it does destroy all!!!! lol...and i have a lot of diehard intel freinds....lol...i already converted two of them to Ryzen..lol...but i want to convert all of them ...lmao... AMD is back to the top and never had any doubt that someday they would be !....
    yeah your ramblibings are suprising making alot of secnce lol
    i've got 1001 working. i have a clock speed of 4.09 gig 100 multi 40.75 multiplayer cor voltage in windows is saying 1.41 SOC is 1.09 ( may be a tad high) ram timings are 16 18 18 18 36 atm i should start to bring down soc voltage a bit more
    i'm running stable at present on this.

    i think i should lower SOC voltage a little more.

  3. #53
    ROG Enthusiast Array SinartNZ PC Specs
    SinartNZ PC Specs
    MotherboardAsus Crosshair VI Hero
    ProcessorAMD Ryzen 7 1800x
    Memory (part number)Corsair CMU16GX4M2C3200C16
    Graphics Card #1EVGA 1080Ti Founders Edition
    MonitorAcer X34
    Storage #1Samsung 960 EVO 1TB NVMe M.2
    CPU CoolerCorsair H110i
    CaseNZXT H440 Black
    Power SupplyEVGA SuperNOVA 750 G2 750W 80+ Gold Full Modular
    Keyboard Logitech G810
    Mouse Logitech G900
    Headset Logitech 633
    OS Windows 10 X64 Pro

    Join Date
    Mar 2017
    Reputation
    10
    Posts
    53

    Yeah this temp issue is starting to drive me nuts... every bit of monitor software I try reports temps into the 70's when under load but when I fire my trusty infrared Thermometer (just had it recalibrated) at it and its not even in the 40's :\

    Quote Originally Posted by muziqaz View Post
    Typical AMD tried to burn their CPUs since Athlon XP days. yes 1.35v will get me long way.
    And fair enough on memory speeds, I guess we need to kick AMDs arse a bit, for them to wake up

    But regarding the temps, I am now running at stock frequencies(turbo is OFF, so 3.6ghz)@1.35v and I have CPU loaded 100% with some AVX code. I locked the fan to 65% top speed which results in 1000rpm. HWinfo reports CPU temps as 75C, which is completely wrong since I can touch easily base of noctua cooler and it is not even warm to the touch. From my previous experiences 75C is very hot to the touch. Contact between CPU and heatsink is perfect: https://flic.kr/p/RU987E

    Is this an effect of that reported temperature offset which AMD for some reason put in? I have no other explanation otherwise.

    Thanks for replies, by the way
    Last edited by SinartNZ; 03-22-2017 at 11:48 PM. Reason: cos spelling and gramar

  4. #54
    ROG Member Array
    Join Date
    Mar 2017
    Reputation
    10
    Posts
    5

    Quote Originally Posted by Demoniacstar View Post
    i do not know about your ram issues ...i do not have that kit so i can't help there...what i can help with is you Display Port issue
    i run three screens two 24" in portrait mod and a 55" 4k display .....i have a HD 60 fps displayport to hdmi adapter ...you can about imagine how hard it is to set bios settings on a portrait mod set up on my 24" monitors .....lol...to get functionality out of your displayport at boot up do this....
    go to bios
    go to boot options
    find CSM (compatibility support module) in that sub menu
    go to launch CSM and set to DISABLED

    that will let your graphics card boot up using the displayport
    Thanks for your answer, but that does not help. Switched CSM off and the Asus Display will just wake up and then Displays no Signal on Displayport ... no boot anyway.

  5. #55
    ROG Member Array
    Join Date
    Apr 2015
    Reputation
    10
    Posts
    18

    Quote Originally Posted by CCarpenter View Post
    Thanks for your answer, but that does not help. Switched CSM off and the Asus Display will just wake up and then Displays no Signal on Displayport ... no boot anyway.
    I have 4k Dell monitor connected to Fury X over display port and had no issues except screen resetting itself to proper resolution and scaling once win 10 boots up. And I have CSM untouched.

  6. #56
    ROG Guru: Yellow Belt Array Demoniacstar PC Specs
    Demoniacstar PC Specs
    Laptop (Model)none
    MotherboardAsus Crosshair VI Hero
    ProcessorRyzen 1800x
    Memory (part number)tdrgd4 16g3000hc16cdc01
    Graphics Card #1XFX R9 390/ EK water block OC'd
    Graphics Card #2XFX R9 390/ EK water block OC'd
    Sound Cardon board
    Monitor2X 24" veiwsonic 1080p/ 55"Samsung 4k curved
    Storage #1Samsung evo 840, 250gig sys drive
    Storage #25 other mixed hdd's
    CPU CoolerEK C6H mono block
    CaseThermaltake core x9
    Power SupplyThermaltake 1500w dps rgb
    Keyboard Corsair k65
    Mouse R.A.T. 9
    Headset Turtle Beach Star Wars X wing edition
    Mouse Pad WOW desk mat
    Headset/Speakers Logitech 5.1
    Accessory #1 Airplex radical 2 480 on gpus swifttech d5
    Accessory #2 Airplex radical 2 480 on cpu swifttech d5
    Accessory #3 Airplex radical 2 280 on sb p500 TT
    Demoniacstar's Avatar
    Join Date
    Nov 2014
    Reputation
    10
    Posts
    160

    Quote Originally Posted by Evanair View Post
    Flashed to 1001.

    Still stable at 41x multiplier and 100 clock to 4.1ghz on CPU @1.4 volts. As for RAM, I was able to decrease my timings from 12-14-14-30 to 12-12-12-30 at 2400mhz @ 1.4 volts.

    T state is still at 2T according to HWinfo.

    Another issue is voltage. Setting RAM to 1.35 volts only gives a reading of around 1.31, upping to 1.4 gives me the 1.35 needed to run the correct voltage through the ram. Ram kit is Corsair 8x2 2400/C10, which I'd like to be able to run at proper timings.
    i think i seen someone say that the cr 1 setting in the 1001 bios did not set cr1 unless your ram wuz at 2666 or higher

  7. #57
    ROG Member Array
    Join Date
    Mar 2017
    Reputation
    10
    Posts
    8

    Cool

    Quote Originally Posted by Raja@ASUS View Post
    Directly from Elmor:

    Alright, BIOS updating time. This version has reverted DRAM settings to 5803 behavior (1T mode), but I also have a test BIOS which keeps 0902 behavior (2T mode) for those who saw improvements there. It also fixes the low temperature reporting and FanXpert OS control.


    1001 (DRAM 1T mode): http://www.mediafire.com/file/1w2qv4...-ASUS-1001.zip

    0038 (DRAM 2T mode): http://www.mediafire.com/file/o5m6ub...-ASUS-0038.zip
    I can`t get stability in Linpack with 1T version, but it`s stable with 2T on HyperX 2666 and 15-16-17-17-32 timings. Also it`s a problem with voltage and cpu temperature. It is overestimated by 0.044V on cpu and ram: setting cpu V to 1.35 = 1.4 real indicated in bios and AI Suite. The same with RAM V: 1.2 = 1.244. Temperature of OC 1700x@3.900 is 76-80 deg., but i know what`s 70 deg. on FX-8320 - fire flames behind of PC. So it`s not 76-80, max 60 deg. Hope, theese and some another bugs will be fixed
    Last edited by cauter; 03-23-2017 at 01:09 PM.

  8. #58
    ROG Enthusiast Array
    Join Date
    Mar 2017
    Reputation
    10
    Posts
    37

    Some quick results here:

    - Despite being listed in the QVL and here in post 1 the G.Skill 2x8 GB F4-3200C14D-16GVK (Ripjaw V) do not work with the CH6 at 3200-CL14, neither using BIOS 1001 (1T) nor 0038 (2T). Both result in a "driver_corrupted_expool" Bluescreen after less than 15 minutes of running Realbench without CPU overclock. Curiously they ran nearly 2 hours of Memtest without error, but that's worth little when Windows doesn't work stable.

    - Using only one bank (2b) of 8 gb ran 1 hour of Realbench without issues, but resulted in another "driver_corrupted_expool" Bluescreen during normal desktop work/load later on.

    - Using stock clocks of 2133-CL15 caused a error 8 (DRAM LED) at one point, the CPU was running at stock (all auto) settings, too. After a reset I got the F1 message telling me that my "overclock" did not work. There was no overclocking active whatsoever, everything set to optimized defaults.

    - Trying to overclock the CPU and running Prime95 (small FFT to get RAM out of the testing) I noticed that at 4.1 gHz I get error 8 crashes (black screen) with the DRAM led being lit. The interesting part about this is that HWInfo logged "CPU Cache L0" Windows Hardware Errors (WHEA) before the crash happened! So error 8 does not seem to entirely be DRAM related, despite the LED being lit.

    - The CH6 CPU temperature sensor seems a bit sluggish compared to the Tctl sensor. Running that Prime95 test at over 1.4v made tCTL jump over 90
    °C right away, while CH6 CPU temp stayed around 60°C for several seconds. This also causes fans to ramp up more slowly.

    - AMD Ryzen Master reports a slightly different Vcore than HWinfo and CpuZ, which one is correct?

    - AI Suite 3 reports a slightly different BCLK and CPU clock compared to HWinfo and CpuZ, which one is correct?

    - Fancontrols are still not working properly, specifically the calibration and resulting minimum speeds.

    - I tried measuring Vcore via the onboard measuring points (between Vcore and Ground). They are rather hard to reach, especially inside a chassis, but I tried a dozen times and I could not get any readings?!
    Last edited by Timur Born; 03-23-2017 at 07:49 PM.

  9. #59
    ROG Enthusiast Array
    Join Date
    Mar 2017
    Reputation
    10
    Posts
    37

    Another thing of note: When Vcore is set to Auto then it jumps all over the place during mostly *idle* times. I saw values ranging from 0.4v to over 1.5v, the latter of which seems especially wrong. Values differ a bit between AMD Ryzen Master and HWinfo/CpuZ, but the maximum being far too high for a stock (Auto) clocked CPU remains the same.

  10. #60
    ROG Enthusiast Array SinartNZ PC Specs
    SinartNZ PC Specs
    MotherboardAsus Crosshair VI Hero
    ProcessorAMD Ryzen 7 1800x
    Memory (part number)Corsair CMU16GX4M2C3200C16
    Graphics Card #1EVGA 1080Ti Founders Edition
    MonitorAcer X34
    Storage #1Samsung 960 EVO 1TB NVMe M.2
    CPU CoolerCorsair H110i
    CaseNZXT H440 Black
    Power SupplyEVGA SuperNOVA 750 G2 750W 80+ Gold Full Modular
    Keyboard Logitech G810
    Mouse Logitech G900
    Headset Logitech 633
    OS Windows 10 X64 Pro

    Join Date
    Mar 2017
    Reputation
    10
    Posts
    53

    Good know I'm no the only one that has being seeing this rather odd behavior. Vcore in Auto is rather like watching a never ending yo-yo.

    Quote Originally Posted by Timur Born View Post
    Another thing of note: When Vcore is set to Auto then it jumps all over the place during mostly *idle* times. I saw values ranging from 0.4v to over 1.5v, the latter of which seems especially wrong. Values differ a bit between AMD Ryzen Master and HWinfo/CpuZ, but the maximum being far too high for a stock (Auto) clocked CPU remains the same.

Page 6 of 268 FirstFirst ... 4 5 6 7 8 16 56 106 ... LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •