Jump to content
HWBOT Community Forums

pepinorang

Members
  • Posts

    358
  • Joined

  • Last visited

  • Days Won

    1

Posts posted by pepinorang

  1. Ok, so the competition is over. For those who wonder how to OC i3 on LN2 on XPower there are a few tricks. First of all is to set a good base profile which you can boot at around -120C. Set both mem and CPU as high as you can. Look down timings on memory, and especially RTL and IOL. Set RTL a little higher or you will reach memory limit quickly while overclocking in OS. I used 52/53 RTL and the corresponding 8/8 IOL for 2000+ memory. Once you have booted your base profile at -120, increase bclk, go down to -170 or whatever your board/CPU can handle and reboot. The key is to not change settings so much that the board needs to do full reset. A "light" reset should be able to boot at around -170.

     

    Next thing to consider is memory on XPower. XPower eats SA-volts for breakfast. Mine scaled to 1.50V. Also, like I said before, set higher RTL/IOL than on 2 dimm boards or you will be limited at around 2000 C12.

     

    If you suffer from low efficiency in Geekbench, remember to turn on AES instructions, as they are off by default on MSI XOC bioses.

    Nice result and thanks for the feedback man! RTL/IOL is dependant of IMC, same for required SA. You can check page 25, I posted XTU @2047 12 12 12 tight, RTL 50/52 with not more than 1.35v SA on a "random" 6.5G Caseking CPU. ;)

  2. Guys, I'm trying instal XP with this guide on XPOWER, no matter what I do, I always get dash blinking at left top corner after step 2 and nothing futher...

     

    I tried so many images, and approachs, even the one from ASUS Z170 thread.

    @pepinorang I saw you mentioned incoming XP instal guide, is there any? thanks

     

    Also, I'm trying to do 4k tight on cold but I get 55 loop no matter what, I tried 90BCLK and increse it from OS to get 4k, but no luck. I'm trying it because I saw many mentions, that its possible on cold, even boot straight with 4k tight

    I think the answer for the upcoming guide is clear from my last post already.

  3. @pepinorang

    We need XOC bios for i3 on XPower. The 173 can't do full pot because PLL volts are limited at 2.0. The GOC bios won't go higher than 125 bclk.

    As some of you already know, I'm no longer in charge of OC community and support for MSI as my job position changed. For any OC requests please contact @MTC. BTW, this is the last XOC competition (online or offline) from me and overall from MSI as well.

     

    Envoyé de mon ASUS_Z010D en utilisant Tapatalk

  4. "- VRM is not suitable for XOC but it is 100% enough for the current comp."

    VRM suitable/not suitable for Stage 3 GB3 All Out ? Oh right, not for Stage 3 as you will most likely be limited with only 3 CPU phases.

     

    "WARNING: Post at 12 11 11 takes FOREVER!"

    It does on all boards I've tried

     

    "Hardmod guide for SA: "

    No hardmod for MPower VRM board, lol. This would be killer for Comp

    We have thought about it before Computex but you would need a G-POWER since vCore+VCCSA+IGP must all start at same time and also timing is very crucial otherwise CPU won't initiate... Probably not worth the trouble.

     

    BTW, since there is no ACPI BIOS, you will notice that performance in 32M vary a lot from run to run (up to 4s).

    Unless you find your own solution, my advice is to re-run again and again, according to BenchBros 2/10 runs are OK: BenchBros`s SuperPi - 32M score: 4min 31sec 562ms with a Core i7 6700K (NO ACPI RUN)

     

     

    [highlight]HAVE FUN![/highlight] :D

  5. No ACPI BIOS for little board..?

    How to get XP working on the board..?

     

    Sent from my Nexus 5X using Tapatalk

     

    @MTC ?

     

    - No ACPI BIOS and there won't be.

    - Use ASUS Guide and press F5 to select MPS Multiprocessor at same time you have to press F6 to load FiraDisk Driver, this will install non-ACPI XP.

    - VRM is not suitable for XOC but it is 100% enough for the current comp.

    - Z170I GAMING PRO AC is an entry level GAMING motherboard and therefore it is much less convenient to OC than XPOWER as it don't have any buttons, debug led or clockgen (need to Clear CMOS/Discharge MB after each memory training fail to reset Intel ME otherwise BCLK stay locked), however it can run memory at higher frequencies and tighter timings.

    - Command Center Lite 2.0.0.11 to 2.0.013 (latest) can be used, no need for special one.

     

    Latest BETA BIOS: https://www.dropbox.com/s/h28ewvqmbyox2cj/E7980IMS.181?dl=0 (I recommend to use this one unless you need more SA)

    Old XOC BIOS (Computex 2016): https://www.dropbox.com/s/dc9pmdwxr81koo1/E7980O17.T72?dl=0 (Only difference is max SA air ~1.33v / max SA Ln2 ~1.35v, I wouldn't use it unless you need more than 1.25v SA which shouldn't be the case for benching. WARNING: Post at 12 11 11 takes FOREVER! 12 12 12 is OK)

    Hardmod guide for SA: https://www.dropbox.com/s/0fjr7oy1xyydyz3/Z170I_gaming_pro_AC.docx?dl=0 (Need to use XOC BIOS, up to 1.8V after vmod)

  6. "1.35v SA is not enough to damage CPU, as once again it is within Intel specs.

    Enhanced Memory Controler auto adjust OC socket voltages to optimize memory OC stability at high frequency, there is no way it can't damage memory.

    What was the post code of the dead CPU? Easy to find out if it was IMC related if you can remind it.

    Anyway, when I kill a CPU that I have oce'd I don't blame the motherboard, I blame myself. However I'm pushing way more than you and for a much longer time but I have yet to kill any 5960x on X99 with +200pcs tested."

     

    Yer, well i am just not sure how it happened, as it was at auto volts and i dont have the cooling power to do anything outrageous anyway, just the 4.5 core, 4.0 ring, 3200 ram benchmarks. It would function fine with ordinary daily use. Never tortured it with encoding or priming.

     

    Diagnostics:

    If i installed 1 stick of RAM, it would turn on then off in about .5 seconds.

    If i installed 4 sticks of RAM, it would hang at debug code 60 and stay powered on.

     

    There might be a motherboard fault with power delivery.

    I have had other boards that dies (Asrock X99 Extreme11 that died about 7 days after purchase)

    Maybe the PSU is surging power (AX1200i) as i have had this for 3 years now and my previous setup died in the same manner with debug code 60 (3930K)- but that is out of warranty.

     

    Seems like IMC indeed, what were the volts for RING?

     

    sabishiihito, now you're are talking my language!! I was considering selling my CPU to a friend but couldn't find enough reason to do so since the IPC was basically the same on Broadwell-E vs Haswell-E. But if I get more options for a memory divider that is good enough for me (I have a hard time getting rid of my upgrade itch).

     

    Can anyone tell me what the max divider for Broadwell-E on this mobo is at 100mhz BCLK?

     

    Max ratio is DDR4-4000 but for 4 DIMM OC, 3500+ is possible on air with a very good IMC, and they are harder to find than 5960X.

  7. Like Leeghoofd said, Computex did take all my time and it is not over yet. :(

    Also busy with testing BD-E and writing OC guide for X99.

    I'll update the first page with new models/BIOS/tools asap but for sure not this week.

     

    Yeah sadly and good to see i am not the only one. But dont talk about that here because its not possible that it can be the Godlike whos destroying ur ram, its the cheap ICs which memory vendors using (just kidding).

     

    And yes u have to get or rma ur kit cuz one of ur sticks is defective now.

     

    I got it that with OCed high end ram there is more likely a fail to happen. But so much people have the same problem on the Godlike that its not just that anymore. Its so fishy....

     

    Still waiting all the info I asked so much times from you to investigate your case! :rolleyes:

    Including all those forums links of "so much people"with similar issues, because until now it is only you + 2 peoples for thousands boards sold.

     

    It can't be cheap ICs! The Godlike board has already destroyed one G.Skill TridentZ DDR4 stick and TWO Corsair Dominator PLATINUM DDR4 sticks of mine. Are you telling me two of the best, well known, and top quality RAM vendors like Corsair (Dominator Platinum Line) and G.Skill (TridentZ line) both their top tier brand of memory are both cheap ICs? No. I can't believe that. I just can't.

     

    Besides, I have ALWAYS been honest and supported the Godlike motherboard (I even just bought a Godlike Carbon board, so let's hope that one doesn't destroy my memory chips like my original Godlike). Pepi and I are on good terms so I hope he sees that we're NOT lying about this issue. I have a TON of respect for Pepi here and I just hope he sees that it COULD have been a board issue and not a RAM issue when two different top tiers of two top quality RAM sellers have been destroyed on the BIOS updates.

     

    Finally someone I can talk to! ;)

    Like I said to you before GODLIKE GAMING and GODLIKE GAMING CARBON are the same boards. Actually you can even crossflash BIOS with a simple trick but shuuuush! :D

    I just tried to flash BIOS 5x times (5 different boards) from v1.2 to v1.4 with the same CORSAIR memory kit you use, nothing happened.

    I will let memtest run overnight and try again to flash with moderate OC settings tomorrow.

     

    If on my Rampage V Extreme I could get 44 ring ratio at 1.246v you thinking I'm stretching it to get 40 ring ratio at 1.2v on a non Turbo motherboard?

     

    Here's hoping, should get my X99A Sli Plus in a day or two, it's been shipped. :)

     

    Without Turbo Socket, Ring OC is limited to 3.7G-3.8G max with LN2 but I can't remind what it can do on air/water, sorry (it might be the same though).

     

    Why I always like ASUS bios updates. You can update a BIOS using Flashback with no CPU, RAM, nothing in the board.

     

    I might just put one stick in for BIOS updates and hope that minimizes the risk.

     

    We also have "BIOS Flashback" feature on Z170A GAMING M7/M9/XPOWER, however it wouldn't be useful here and not very practical to remove components if you are running full system in case... :D

     

    Well,

    My 5960X died, after 10 months of use, never abused it.

    Didn't do anything stupid to it either, just auto volts.

    It was replaced by intel but still cant understand why it would die.

     

    This is what happened:

    Using PC, then did shutdown from windows.

    Went out for 2 hours of errands.

    Came home, pressed power button and no display.

     

    Next day i went to the PC shop and we were using the process of elimination.

    I thought DIMM slot 1 was dead on the board, so after swapping out rams and cpus, ended up being the CPU.

     

    Still want to get close the 6900K OC speeds, but what the hell would of killed it with auto volts?

     

    Guessing it could of been constant heat as the thermal paste wasn't very good, it was very cured and heavy to wipe off. Using different paste now and it runs at lower temp.

     

    I even had the internal vr ovp ocp protection enabled.

     

    *Updated to 1.4 with a default bios reset first, and all 4 ram sticks in. Ram is ok after bios pdate.

     

    ** i have a feeling the Enhanced Memory Controller setting may have done something to the dead cpu and maybe other users memory faults.

     

    *Could the memory at 3200Mhz be harming the cpu by setting the SA to V1.350 on auto config?

     

    ** Never ran prime95 with 16 threads / or in a tortureous manner to the point where it was killing it, just quick benches with Realbench or Cinebench or X264, where it would run just 20 mins or a run through and screenshot the results.

    Using the system as a workstation, not a bench station.

     

    1.35v SA is not enough to damage CPU, as once again it is within Intel specs.

    Enhanced Memory Controler auto adjust OC socket voltages to optimize memory OC stability at high frequency, there is no way it can't damage memory.

    What was the post code of the dead CPU? Easy to find out if it was IMC related if you can remind it. However, I have been pushing 24/7 and XOC for a much longer time but I have yet to kill any 5960x on X99 with +200pcs tested.

     

    To everyone, FYI before releasing a product and MP BIOS it is not only tested by our BIOS team, QVL team running memtest overnight but also by our DQA team that test devices compatibility and also run burning tests overnight (last step), also most of the memory vendors do test those BIOS. It is not like some random programmer code the BIOS and post it on the website. So it is OK to report bugs and provide feedback negative or positive, but unless you can prove such dramatic claims/assumptions with factual data I suggest you keep them for yourself. Thank you very much.

  8. First page has been updated with new XOC BIOS + new flashing tool (as usual XOC BIOS can't be flashed with M-FLASH). :celebration:

     

    Change log:

    - Fix PCI-E lock to 1.1

    - Fix B-DIE need to boot 3866x99~99.8

    - Fix E-DIE need to boot 4000x99~99.8

    - Replace old Micron/Hynix memory presets with new ones for AFR/MFR/D-DIE/E-DIE/B-DIE (I'll add Hynix M-DIE and update B-DIE asap, it only set primaries and tFAW for the moment)

    - Add Non-K OC support

    - Add ACPI support (need to be tested, not sure it works properly on this one)

    - Increase SA/IO voltages in OC profiles

    - Rename LN2 LOW OC profile to LN2 -120

    - Add "LN2 FULL POT" OC profile

     

    [highlight]***This BIOS must be used with Command Center Lite HOT version***[/highlight]

     

    New memory presets:

    attachment.php?attachmentid=4308&stc=1&d=1463129307

     

    Non-K OC + ACPI:

    attachment.php?attachmentid=4306&stc=1&d=1463129307

     

    New OC Profiles:

    attachment.php?attachmentid=4307&stc=1&d=1463129307

     

    HAVE FUN GUYS! :D

  9. I am very impressed.

    I must put my hands on one of these beauties.

     

    Good work !

     

    Even if you have probably noticed, I must warn that I needed -80 on IMC for +2040 (memory weren't cold though).

    The CPU I used isn't binned in any way so maybe a better IMC could also do on air.

    However for XTU @2000 12-12-12-28-220-1T tight, ALL AIR is fine:

    attachment.php?attachmentid=4305&stc=1&d=1463124978

     

    I'll try 32M asap, but for now I can't get this new BIOS to work with XP (ACPI or not).

    At least it pass without wazza in 7 x64 :)

     

    For XP installation you must press F7 when the System asked you for the AHCI Driver

     

    It is not the issue, I had XP running without ACPI before.

    This BIOS is supposed to be ACPI compatible (I saw it working on another version before) but neither my old install without ACPI or a new install with ACPI are working...

    Weird but I have no time to check this further at the moment it'll have to wait a bit.

×
×
  • Create New...