Please report all spam threads, posts and suspicious members. We receive spam notifications and will take immediate action!
Page 1 of 2 12 LastLast
Results 1 to 10 of 14

Thread: Have same problem with my second X58A-UD3R 2.0 build




  1. #1
    Join Date
    Sep 2010
    Location
    California
    Posts
    262

    Default Have same problem with my second X58A-UD3R 2.0 build

    My system:

    i7 950
    Mugen 2
    X58A-UD3R 2.0
    Corsair 6x2GB 9-9-9-24 1600 1.6V HX3X12G1600C9 G
    Newegg.com - CORSAIR XMS3 12GB (6 x 2GB) 240-Pin DDR3 SDRAM DDR3 1600 (PC3 12800) Desktop Memory Model HX3X12G1600C9 G

    BIOS:
    CPU Clock: 16 / want 19
    QPI: 36
    Uncore: 18
    BCLK: 200
    Mem Multi: 8

    9-9-9-24

    VCORE: 1.3
    QPI/VTT: 1.335
    CPU/ PLL: 1.8

    IOH CORE: 1.1
    ICH CORE: 1.1
    VDIMM: 1.6

    What I want: to raise CPU Clock to 19

    What happens if I raise CPU Clock to 19?:

    This is my new system. I have another system that is the same except for its i7 930. The regulars here at Tweaktown helped me guess how to make my system run properly. But we never figured out what the problem actually was.

    With the i7 930 system running stable 24/7, I felt comfortable buying a similar system for my second computer. But I was surprised to see that the i7 950 gives different V options in the BIOS. For instance, because of my 6dimm memory set, my first QPI/VTT runs at 1.4. But the i7 950 has to run at 1.335 because there is no 1.4.

    When I first put my new system together, with the CPU Clock at 19, MEMTEST would usually fail at one spot. But occassionally, it would pass the test. And whenever it passed the test, the Intel Burn test set to maximum and 20 rounds would also pass.

    But after several Intel Burn tests, it stopped booting with CPU Clock at 19. Now, it will only boot with CPU Clock at 16.

    What could this mean?

    One thing that helped me fix the problem the last time was reinstalling the Mugen 2. It turned out that I had accidentally scraped away part of the thermal paste during installation. But my new system is a Lian Li PC-7F, which makes it easier to install the Mugen 2.

    I really don't want to reinstall the Mugen 2. Besides, the CPU temps are very low.

    So, do you fellas have any insights for me? ^^;

  2. #2
    flyfisher is offline Member
    Join Date
    Dec 2008
    Location
    South Central PA
    Posts
    191

    Default Re: Have same problem with my second X58A-UD3R 2.0 build

    Looking at your BIOS settings I would first try setting uncore back to 16x. 3600 Mhz uncore takes a lot of QPI volts, 1.335 isn't enough, especially if you are trying to run 6 sticks of RAM. At 16x the uncore is a manageable 3200 Mhz.

  3. #3
    Join Date
    Sep 2010
    Location
    California
    Posts
    262

    Default Re: Have same problem with my second X58A-UD3R 2.0 build

    Quote Originally Posted by flyfisher View Post
    Looking at your BIOS settings I would first try setting uncore back to 16x. 3600 Mhz uncore takes a lot of QPI volts, 1.335 isn't enough, especially if you are trying to run 6 sticks of RAM. At 16x the uncore is a manageable 3200 Mhz.
    Hello, flyfisher. The Uncore did work at 18 at times before. But now, with CPU at 19, the Uncore does not work at 16, 17, or 18. At the moment, with CPU at 16, the Uncore works at 18.

    Something is very odd about my system. If Memtest fails, then no matter how I change Vs, it will always fail if I do not power down first. I have to turn the system off and not just reset.

    So, I do not know that the QPI/VTT or the CPU Clock is the problem.

    Right now, the QPI/VTT safe limit on my BIOS is 1.335. The next step takes me over 1.35, which is colored magenta in the BIOS and that does not fix the problem either.

    :'(

  4. #4
    flyfisher is offline Member
    Join Date
    Dec 2008
    Location
    South Central PA
    Posts
    191

    Default Re: Have same problem with my second X58A-UD3R 2.0 build

    Well that is odd. Uncore and CPU multiplier really should have nothing to do with each other. As you probably know the uncore must be at least 2x the memory, that's why I suggested 16x for uncore. Before you first booted the machine, did you clear CMOS and load optimized defaults? Does your Corsair memory have an XMP profile? If it does I would load that since the profile would set the timings for 1600 MHz.

  5. #5
    Join Date
    Aug 2009
    Location
    Birmingham,Great Britain
    Posts
    9,250

    Default Re: Have same problem with my second X58A-UD3R 2.0 build

    what do you mean no 1.4v for qpi/vtt? 1.335 isnt going to be enough for 12gb's @1600mhz,you should be able to run at 21x cpu multi or 19x,have you tried underclocking the ram slightly? you can use 1.3-1.45 qpi/vtt voltage even 1.5v but you shouldnt need that much
    Gigabyte z77x UP4-TH F11c Modded Bios
    Intel i7 3770k 24/7@4.8ghz 1.38v Turbo llc +0.165v dvid multithreading enabled
    Samsung Green(MV-3V4G3D/US) 8GB @2133mhz 9-10-10-21-1t 1.55v
    Thermalright Silver Arrow Cpu Cooler
    1xSamsung 840 pro 256 Gb SSD windows 8.1 pro 64bit
    1xSamsung f4 HD204UI 2tb hard drive Storage
    Powercolor 7970 3gb V3 @1150mhz core/1700mhz mem,1.150v Accelero aftermarket air cooler 55c max
    Razer Lycosa Keyboard
    Logitech X-530 5.1 Speakers
    Lite-On iHAS124-19 24x Sata DVDRW
    K-World Hybrid DVB-T 210SE Digital T.V Card
    L.G E2260V L.E.D 1920x1080 Monitor
    Xfx Pro 750w silver rated Psu 80+
    Fractal Arc Midi Case

    http://i38.tinypic.com/14myvfa.jpg x58 ud5 <=3.8ghz + 4.2ghz Overclock Template!!
    http://www.youtube.com/user/warren304#p/u Visit Me On Youtube

    Lots Of Gaming Videos With X58 Ud5 System And Gpu On My Youtube Channel!!
    Just Uploaded New Battlefield 4 Video!!!!!!!!!!!!!!!!!!!!!!

  6. #6
    Join Date
    Sep 2010
    Location
    California
    Posts
    262

    Default Re: Have same problem with my second X58A-UD3R 2.0 build

    "Well that is odd."
    --flyfisher

    Unbelievable!!! >:(

    I put the computer together on Sunday. And by Tuesday, it became unusable. Fortunately, I recognized the pattern. Progressively, I was forced to cut back on the overclock and Vs until I could not even run the system underclocked.

    But that is the exact same thing that happened on my first system!

    Last time, it turned out that when I took off the Mugen 2, there was a gap devoid of thermal paste running vertically down the center of the chip. The exact same thing happened on this build. When I took apart the Mugen 2 on this build, there was that same vertical gap. It is about 5mm across and runs down the center of the chip--exactly where I laid down the thermal paste.

    I do not know how I keep scraping the paste, especially since, minus that gap, the rest of the surface was completely covered.

    This time, after I shoved the Mugen 2 on the chip, I pulled it off again to see the paste coverage. It was what I expected--the paste covered a circular area around the center of the chip, leaving the rest of the surface uncovered by the paste.

    Anyway, I will know if the scrape has somehow returned after I spend tonight and tomorrow running the Intel burn test. If Memtest identifies memory corruption immediately after the passing the test on maimum, I will know that the scrape has somehow returned.

    ...

    "what do you mean no 1.4v for qpi/vtt? 1.335 isnt going to be enough for 12gb's @1600mhz"
    --wazza300

    Yep, that is correct, the X58A-UD3R + i7 950 gives a different set of V options than with an i7 930. I was surprised, too. But the i7 950 really is superior to the i7 930. I am currently running with CPU at 19 with VCORE at 1.3. And my 6x2GB set runs at 9-9-9-24 1600 at 1.6 VDIMM and 1.335 QPI/VTT. And the chip actually runs several degC cooler in a case that has an inferior air cooling system than my other case. I wish the i7 950 was as cheap as it is today when I bought the i7 930 several months ago.

    ...

    Like I said above, I hope the Intel Burn test confirms that the thermal paste has not somehow been scraped down the center again. Unbelievable. I installed the Mugen 2 four times now on only two systems. And in two of those times, that gap appeared.

  7. #7
    Join Date
    Sep 2010
    Location
    California
    Posts
    262

    Default Re: Have same problem with my second X58A-UD3R 2.0 build

    Ai.

    It did not work.

    As I expected, my system with the mugen 2 reinstalled passed memtest with the overclock on the first try. And then it passed the Intel burn test.

    But when I ran memtest again, it failed.

    I take this to mean that something has gone wrong with the thermal paste again.

    So here is the question. I am running out of paste and cleaning solutions. Do I really have to wipe off the old paste before I reinstall the mugen 2? I need to see what the paste coverage looks like. So I am going to pull the hsf off again.

    If the vertical gap has somehow returned, I just want to dab some more paste there.

    Also, how do you install the hsf properly? I wonder if I am screwing it on too tightly.Should I leave the hsf a bit lax?

  8. #8
    Join Date
    Oct 2010
    Location
    Newcastle, Australia
    Posts
    59

    Default Re: Have same problem with my second X58A-UD3R 2.0 build

    so you are getting a vertical stripe of no paste after a period of time while using a mugen 2, do you have another cpu cooler you can try?

    with your answers to the qpi/vtt are you sure your looking at the correct feature in the bios?

    i think you should get your pc running at normal speeds before atteempting to oc it.
    Mobo - GA-X58A-UD3R v2 bios version FB
    Ram - Mushkin 998826 ridgeback-(3x2gb) 6gbrm 6-8-6-24 1T 1.65v Qpi-1.355v
    Cpu - i7 950 3.07mhz
    Cooling - Corsair h70
    Gpu - 2 x evga nividia gtx 460 1gb external exhaust models
    Psu - Corsair hx850
    Hdd - 1tb western digital black caviar
    Case - Antec nine hundred and two
    Os - Windows 7 home premium 64bit

  9. #9
    Join Date
    Sep 2010
    Location
    California
    Posts
    262

    Default Re: Have same problem with my second X58A-UD3R 2.0 build

    "so you are getting a vertical stripe of no paste after a period of time while using a mugen 2"

    Not after a time. I think it happens during installation. I am not sure how. My first system is running fine overclocked with the Mugen 2 now.

    "do you have another cpu cooler you can try?"

    No. I left the Intel cooler at another place.

    "with your answers to the qpi/vtt are you sure your looking at the correct feature in the bios?"

    Yes. The QPI/VTT option goes: ...1.315, 1.335, 1.355, 1.375,....

    "i think you should get your pc running at normal speeds before atteempting to oc it."

    What's bothersome though is that if the problem is the paste, over time, the system may not be handle the V. A few rounds of Intel Burn tests set to max memory and 20 seems to make even underclocking progressively worse if the pste is not applied properly.

    In other words, if the system was actually stable with the overclock and then is not some time soonafter, doesn't that seem worying? When the system completely failed on my first attempts on both my systems, I ended up unable to run the computer even severely underclocked.


    I am running tests with CPU Clock on 16 now. But from experience, I think that this particular system really should not have difficulty running at 19.

    So, do you think it is not too late to apply osome more paste if I remove the cooler at this time...

  10. #10
    Join Date
    Sep 2010
    Location
    California
    Posts
    262

    Default Re: Have same problem with my second X58A-UD3R 2.0 build

    My i7 930 needs only 1.34V for QPI/VTT. I do not think it is a good idea or necessary to go over that V with this ram set-cpu-board combination running 9-9-9-24 1600.

Page 1 of 2 12 LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

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