Page 5 of 7 FirstFirst ... 3 4 5 6 7 LastLast
Results 41 to 50 of 63
  1. #41
    ROG Member Array
    Join Date
    Sep 2012
    Reputation
    10
    Posts
    18

    Still nothing.......... wow raja why?????

  2. #42
    ROG Enthusiast Array
    Join Date
    Sep 2012
    Reputation
    10
    Posts
    59

    awww pooo!!

    Just when I went and bought some 2400 ram too!!
    Will do some tests with my current 1866 before I change over to using the 2400 stuff (G Skill Ripjaws 4x4Gb 1866 C9 to G Skill Trident 4x4Gb 2400 C10)
    1700X B350/RX570 STRIX

  3. #43
    ROG Guru: Yellow Belt Array
    Join Date
    Sep 2013
    Reputation
    10
    Posts
    162

    the Original X79 is almost four years old, that's why!
    Raja wont even comment cause then its like they are saying or admitting there is a problem............

  4. #44
    ROG Enthusiast Array extreme's Avatar
    Join Date
    Jan 2011
    Reputation
    10
    Posts
    54

    But they still updating the Bioses for x79 platform without any considration or fixing. ho what a waste.
    if they dont believe there is a problem why dont they do some testing in their labs just to confirm.

  5. #45
    ROG Enthusiast Array
    Join Date
    Sep 2012
    Reputation
    10
    Posts
    59

    Can confirm I have the same issue.
    Using 4x4Gb 1866 C9 - I get 52k read and write speed
    Using 4x4Gb 2400 C10 - I get 58K read but only 42k write



    Look at that abysmal write speed on the 2400 ram

    Last edited by zoomee; 07-19-2014 at 05:43 PM.
    1700X B350/RX570 STRIX

  6. #46
    ROG Enthusiast Array
    Join Date
    Sep 2012
    Reputation
    10
    Posts
    59

    A friend of mine tested this on his 4930k + RIVBE. He got the same results using the 100 strap and 2400Mhz memory.

    What was also interesting was that he got the same poor write results whilst using the 100 strap and 2133 memory!

    1700X B350/RX570 STRIX

  7. #47
    Derp Array
    Join Date
    Sep 2013
    Reputation
    10
    Posts
    20

    Moved from a RIVE with a 3970x to a RIVE BE with a 4960x - using the same ram - CMD16GX3M4A2400C9 ( Corsair Dom Plats 2400mhz CL9 )

    write speeds in AIDA64 on the 3970x with the RIVE were in the 70k mark , the RIVE BE with the 4960x i get MAX 43k no matter what! - posted a thread a few days ago with my pics here ..

    http://rog.asus.com/forum/showthread...347#post423347


    Running the latest UEFI :/// I just spent all that to get worse performance outta newer tech ?


    EDIT: Just did the clock / ratio "fix" mentioned on page one and my write speeds and mem latency jumped back to what it's supposed to be .... I really don't wanna do this everything my machine posts ://

    Before Pic

    Click image for larger version. 

Name:	4960x.jpg 
Views:	25 
Size:	101.9 KB 
ID:	39594

    After - sorry I used 45 multi here, but the restults are the same ^^

    Click image for larger version. 

Name:	4960x_2.jpg 
Views:	24 
Size:	102.0 KB 
ID:	39595
    Last edited by Wayâ„¢; 07-30-2014 at 06:07 AM.

  8. #48
    Derp Array
    Join Date
    Sep 2013
    Reputation
    10
    Posts
    20

    Sooooooo apart from the work around - does anyone have anything on this ?

  9. #49
    ROG Guru: Green Belt Array
    Join Date
    Dec 2011
    Reputation
    60
    Posts
    655

    Hopefully ASUS will chime in if this stays near the top.
    Not buying a 2400 kit until they can iron this issue out.
    MB: ASUS Rampage IV Black Edition
    CPU: Intel Xeon E5-2697 v2 (Cooled by Corsair H100i w/ Noctua NF-F12 fans)
    RAM: 64GB G.SKILL RipjawsZ 1600 (10-10-10-30)
    GPU: EVGA GeForce GTX Titan X Hybrid
    TV Tuners: 2 * Hauppauge HVR-1800
    Case: CoolerMaster Cosmos 2


  10. #50
    ROG Enthusiast Array
    Join Date
    Apr 2014
    Reputation
    10
    Posts
    31

    I'm on 0701 right now, haven't tried the new one yet, I don't see the point actually, I'll probably wait it out for the next ver, even if there isn't a fix in that one.

    I should install linux again and check with that, but in windows, not touching the power management cpl (say forma fresh install), with the bug the cpu never goes over the stock cpu multi.
    Without the bug it climbs to where the bios has set the cpu multi.
    Mafia 2's benchmark is where I noticed the prob with that, when i 1st got the board I had poor perf in that bench, eventually got rid of the prob but I wasn't really sure how at the time (fiddled with powermanagement in windows...).

    When I recently did a fresh install of windows again (I tweak alot and mess up the registry alot too lol), I had the poor perf in mafia 2 again and I was like wth...
    For the heck of it, did a aida64 bench and realized, hey windows is acting like linux, cpu is stuck at 32x.
    Did the work around for the mem bandwith and surprise the cpu multi bug was fixed without even touching the power management in any way.
    ^^

    So the underling issue is more then just a mem div issue.

    Take this example, you buy a rig, get either a k or x cpu.
    And you don't touch windows power management, or perhaps you use linux...
    Then eventually you realize, hey your cpu isn't overclocking at all, you bought thee highest end overclocking board and cpu and you can't overclock it...

    The bug is pretty easy to work arund once you learn how, but it's very annoying.
    However, my issue is the bug at higher bclk, 100mhz strap.
    Original bios I had no probs getting the workaround working.
    Now I can't get the work around working at all at higher bclks.
    I want my 114mhz...
    Easy to get it working at 100mhz, but clocking up the bclk, it'll eventually need a power cycle to re-setup the hidden thingys to post.
    110 and 114mhz is to high to jump from 100mhz without a power cycle.
    (Issue is, the work around doesn't work at higher bclks anymore lol, so I try the work around at 100, then try posting at 110 without a power cycle, not happening lol..)

    Using a diff strap, and I lose the ability to idle my cpu.
    Ok fine for peeps that want suicide runs, but I don't care about that, I want a 24/7 overclock.
    So to put it simply, I cannot overclock this board at all with this bug.
    With the bug, bandwith blows, and the cpu can't clock up in linux.

    So I'm stuck waiting, doing absolutely nothing with the rig except for generic use, I don't even feel like playing games with it (I need a smd resistor for my 2nd vga card anyways...).
    Apparently so are a few other people's.
    And I bet there's a whole lot more affected by it and they don't even realize it.
    I don't think there is a single review out there that noticed it, or mentioned it anyways.

    I'm half wondering should I go around the net telling reviewers about this so we can get this stuff plastered all over the tech news sites?
    Like the 290x probs for ex, wasn't halfway fixed until reviewers forced amd's hands and they eventually released a bios...

    How would this look?, and don't take any offense, I'm not actually seriously thinking of doing this right now...
    Reviewers all over start saying: most expensive board, poorest performing board, does not overclock.... ^^ you get the idea.

    Or perhaps it's intel's fault (Actually probably is...).
    We could say, intel's latest cpu generation, performs at 50% of the previous gen and cannot overclock in linux... as an ex..., tagline could be worse lol.
    (Edit: just thinking abut that last line I wrote, sound like a lawsuit against microsoft sorta thing lol, just chill though I'm not threatening or anything, just saying it looks bad because it really is...pretty bad)
    Last edited by NEOAethyr; 08-01-2014 at 06:36 PM.

Page 5 of 7 FirstFirst ... 3 4 5 6 7 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
  •