Jump to content
HWBOT Community Forums

pepinorang

Members
  • Posts

    251
  • Joined

  • Last visited

  • Days Won

    1

Posts posted by pepinorang

  1. I just wanna say. I dislike this way of doing the bios very much! I don't have spare usb stivcks laying around to sacrifice for just one tiny program. I have so many bios from various board on my usb sticks (ones I dont are OS installation) this is really excessive. I mean on asrock i just extract to usb stick and press f5 at the bios entry....

     

    as of now I stil dont have it updated... i opened that program and I just get errors and it exits when I press something

    Test BIOS for XOC have ALWAYS been this way and it will not change. There is nothing complicated to flash with FPT. If you don't like it, you don't use it and stick to MP/BETA BIOS. ;)

  2. First page updated with Command Center Lite Memory 1.0.0.14:

    - Fix several reading bugs, I have quickly checked it, seems OK but please report here if you find any issues.

     

    Long awaited XOC BIOS E7968IMS.HOT along with its specific Command Center Lite! :D

    - HOT BIOS is based on MP BIOS 143, both BIOS will work the same with other ICs than AFR.

    - You need FPT 1194 to flash this BIOS (see first page).

    - ALL voltages fully unlocked for XOC.

    - BIOS include new AIR EXTREME and LOW LN2 OC Profiles which preset CPU settings/voltages from Ln2 OC Guide along with other settings for XOC.

    - BIOS has been fine tuned for Hynix AFR with new PCB (HOT Edition, will be provided by HyperX to contestants).

    - Tested with XTU and AFR new PCB up to 1986MHz 11-16-16-28-300-1T with "loose tertiary timings" on Ln2 (1754pts @5Ghz Core/Ring and 3866 so far, I'll try more later).

    - Tested with XTU and AFR new PCB up to 1900MHz 13-17-17-28-310-1T with "tight secondary timings" on air.

    - If you use AFR with current PCB, maybe it needs some manual fine tune of the option "Memory layout finetune" in "Advanced DRAM settings".

    Start from 1KHz and test up to 2.3KHz with +3733 and tight timings to see which one works the best (I had no time to do it yet, sorry).

  3. Bios 1.41 fixed the reboot issue. I can also confir that 4 dimm now work with 1T. Have also done some testing on 4 dimm vs 2 dimm and there is NO performance gain when pushing frequency and timings with both 2 dimm and 4 dimm.

     

    However I have a problem with my board. It's probably just my board, and problem probably occurred during LN2. But it gets stuck with USB over current protection and shuts off after 15 seconds. I've tried to disconnect all USB devices of course but it doesn't help. Looked at the USB ports but they seem fine. Sometimes I can get the board to boot after resetting bios and powering on like 20 times. Then it works for a while, but after a few reboots it's there again. Is there any way to bypass this? Any way to hardmod it or something?

     

    I saw that before on X99, probably some moisture somewhere.

    Hardmod won't help, Wizerty tried to remove all USB IC and it didn't help on X99A XPOWER AC.

    I think dRweEz told me that drying board in oven did help in its case.

    If not -> RMA

  4. Hi , I tested 1 kit of 4x4 go ripjaws V 3466 mhz c16 on MSI z170A xpower and Asus m8 hero .

    It boots with xmp profile and 1T on Asus board but not on the xpower .(need 2T)

     

    @nvidiaforever2 seems to me that the 1T 4 DIMM issue is fixed in BIOS 141 :)

     

    4 DIMM 1T with manual primaries/All sub AUTO

    attachment.php?attachmentid=3466&stc=1&d=1447412913

     

     

    @Doug2507 this is how much you can gain from fine tuning each timing:

     

    2 DIMM with manual primaries/All sub AUTO

    attachment.php?attachmentid=3465&stc=1&d=1447412902

     

    2 DIMM with almost all tight timings

    attachment.php?attachmentid=3464&stc=1&d=1447412886

  5. Although not directly relevant to Z170 Titanium edition, I would like to offer an update to my current setup of MSI Gaming M5 with four sticks of HyperX 3000 predator.

    After latest bios update, and after some work, I got fully stable under sustained workloads, 3200 - 12-15-15-28-280-1T with pretty tight seconds and thirds with only 1.64Volts !

    ON AIR !

    SA 1.35

    I/O 1.25

     

    Either I have an excellent ram kit ( I think Splave binned forty sticks to achieve

    the same settings ) or a very good board, or both.

     

    I really believe this is great performance from a 200$ board.

     

    In fact a sub 200$ board, official MSRP is 189$ but you can find it even cheaper in the US. :D

  6. the xtu phenomenon 2v4 dimm is to do with the memory trace topology connecting the cpu socket to memory slot, im not sure what the 2 types are called but basically on boards that scale with 4 dimm the memory slots are all connected then a bus goes to the cpu where as on boards that dont scale and perform better with 2 dimm each slot has its own bus to the cpu. its not as simple as just manufacturer for example asrock oc formula performs better with 2 where as the asrock extreme 7+ does scale with extra dimms

     

    What you are referring to is Daisy chain vs T-Topology.

    However your description is not exact, both use two "bus" or memory signals.

    In T-Topology:

    - Signal X goes from CPU than split in 2 to supply channels A1/A2 equally.

    - Signal Y goes from CPU than split in 2 to supply channels B1/B2 equally.

     

    In Daisy chain:

    - Signal X goes from CPU to channel A1 than to channel A2.

    - Signal Y goes from CPU to channel B1 than to channel B2.

     

    I could describe this better with paint but some people might not be happy with this...

    Anyway, both designs have their advantages and disadvantages but to get back to XTU, the "others need 4 DIMM to match OCF" thing is a misconception. :)

  7. Excellent!

     

    Out of curiosity, what's the explanation behind the xtu 2/4 stick phenomenon? I.e, msi needs quad and asrock for example only dual.

     

    MSI don't needs quad, neither ASUS.

    Ask @Strat which spent a lot of time on it and look at his topic for Hynix MFR here: http://forum.hwbot.org/showthread.php?t=144758

    I don't see a lack of efficiency there :)

     

    Also look this score from Hazzan on M8E: http://hwbot.org/submission/3022384_hazzan_xtu_core_i7_6700k_2101_marks vs this XTU score from Splave on OCF: http://hwbot.org/submission/3019044_splave_xtu_core_i7_6700k_2104_marks

    3pts difference, same CPU frequency, both 2 DIMM.

     

    The key is in the timings. ;)

  8. Updated thread with latest BIOS, Ln2 OC Guide v1.4, MSI GAMING CPU-Z v1.74 (new fonts colors, much clearer as requested by many ocers) and Command Center Lite version which now works with all BIOS. :) Lots of updates in OC Guide v1.4:

    - Useful Info, BIOS and Utilities revised

    - Ln2 Tips revised

    - Ln2 Voltage Table revised

    - CPU ST/ST V6 & PLL OC Voltages revised

    - How To Max Out CPU revised

    - How to move CB & CBB lower revised

    - Added BONUS Part 2 with hardmod for CPU ST ;)

     

    For now, I recommend to only use latest MP/BETA BIOS from this thread for XOC.

    EDIT: New Test BIOS based on 141 will be released soon, only differences will be PLL/PLL OC/vDimm unlocked to hardware limit.

    1T issue with 4 DIMM has been solved in BIOS 141.

     

    It is strongly advised to not use JLN1 jumper/short/switch (left side of debug LED, near fan connector) as we recently learned it can cause POST issues on retail CPUs. Support for JLN1 jumper has been removed from BIOS updates after T0P, board won't POST anymore with JLN1 ON.

    • Like 1
×
×
  • Create New...