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

Thread: E7200 + X48T-DQ6 + DDR3 1800MHz = Only up to 3.4GHz with 1.3625V




  1. #1
    Qu4k3r's Avatar
    Qu4k3r is offline Junior Member
    Join Date
    Feb 2009
    Location
    Valencia, VENEZUELA
    Posts
    95

    Default E7200 + X48T-DQ6 + DDR3 1800MHz = Only up to 3.4GHz with 1.3625V

    Hi everybody, I'm new here.

    Last week I recieved a brand new Gigabyte X48T-DQ6 motherboard. My rig is builted with a E7200 and the ram is OCZ DDR3 1800MHz 9-9-9-27@1.9V. Until 5 days ago all were at stock settings. My idea is push it up to 3.6GHz (8x450MHz) with 1:2 (fsb:ram) ratio. Let me show you what I've done step by step and results. Bold text are the changes compared to previous setting. Any comments, help or criticism is always welcome.

    Stock: 266x9.5=2.53GHz, Vcore)=1.2250V, spd(4.00A)=1066MHz, timmings(auto)=9-9-9-24, CR(auto), Vddr(normal)=1.5V ... OK.-

    OC1: 333x9=3GHz, Vcore=1.2250V, spd(4.00B)=1333MHz, timmings(manual)=9-10-10-25, CR(auto), Vddr=1.5V ... UNSTABLE.-

    OC2: 333x9=3GHz, Vcore=1.2500V, spd(4.00B)=1333MHz, timmings(manual)=9-10-10-25, CR(auto), Vddr)=1.5V ... OK.-

    OC3: 388x8=3.1GHz, Vcore=1.2500V, spd(3.33D)=1293MHz, timmings(manual)=9-10-10-25, CR(auto), Vddr=1.5V ... OK.-

    OC4: 400x8=3.2GHz, Vcore=1.2500V, spd(3.33D)=1333MHz, timmings(manual)=9-10-10-25, CR(auto), Vddr=1.5V ... UNSTABLE.-

    OC5: 400x8=3.2GHz, Vcore=1.2750V, spd(3.33D)=1333MHz, timmings(manual)=9-10-10-25, CR(auto), Vddr=1.5V ... OK.-

    OC6: 400x8=3.2GHz, Vcore=1.2750V, spd(4.00D)=1600MHz, timmings(manual)=9-9-9-27, CR(auto), Vddr(+0.35)=1.85V ... UNSTABLE.-

    OC7: 400x8=3.2GHz, Vcore=1.3000V, spd(4.00D)=1600MHz, timmings(manual)=9-9-9-27, CR(auto), Vddr(+0.35)=1.85V ... UNSTABLE.-

    OC8: 400x8=3.2GHz, Vcore=1.3000V, spd(4.00D)=1600MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(+0.35)=1.85V ... OK.-

    OC9: 400x8=3.2GHz, Vcore=1.2750V, spd(4.00D)=1600MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(+0.35)=1.85V ... UNSTABLE.-

    *
    *
    *

    but i want get here

    OC n (Goal): 450x8=3.6GHz, Vcore=???V, spd(4.00D)=1800MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(+0.40V)=1.9V




    According to what I understand, then I must do like this:

    1)Find the lowest Vcore to reach 3.6GHz stable cpu. And memories at lower frec than nominal values (1800MHz), in order to avoid non-stability caused by ram. I must use SPD to for that

    Intel says voltage for this cpu is from 0.85 to 1.3625V. So, I'll set Vcore=1.3625V and go to raise cpu frec to reach 3.6GHz (379x9.5, 400x9) stable. Then, lower Vcore until become unstable. Set Vcore at voltaje to next higher voltage level where it becomes unstable.-

    2)After reaching 3.6GHz, Now is when I must change FSB to 450MHz in order to set ram at1800MHz.

    :::::::::::::::::::::::::::::::::::::::

    Last stable settings:

    OC8: 400x8=3.2GHz, Vcore=1.3000V, spd(4.00D)=1600MHz, latencias(manual)=9-9-9-27, CR=2T, Vddr(+0.35)=1.85V ... OK.-

    ...and a weekend later....

    Now I show you new attempts 3 days after starting from the last stable configuration:

    OC: 413x8=3.3GHz, Vcore=1.3000V, spd(3.33D)=1322MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(normal)=1.5V ... Unstable: Orhtos fails 1rst test.

    OC: 413x8=3.3GHz, Vcore=1.3250V, spd(3.33D)=1322MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(normal)=1.5V ... OK.-

    OC: 413x8=3.3GHz, Vcore=1.3250V, spd(4.00D)=1600MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(+0.4V)=1.9V ... OK.-



    I'm do not want to wait so I decided try 3.6GHz with highest voltage said by Intel: (Vcore=1.3625V).-

    OC: 400x9=3.6GHz, Vcore=1.3625V, spd(4.00D)=1600MHz, ltimmings(manual)=9-9-9-27, CR=2T, Vddr(+0.4V)=1.9V ... Unstable: Fails loading windows or reboots when loaded it.

    OC: 400x9=3.6GHz, Vcore=1.3625V, spd(3.33D)=1333MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(normal)=1.5V ... Unstable: Fails loading windows or reboots when loaded it.

    OC: 400x9=3.6GHz, Vcore=1.3625V, spd(2.66D)=1066MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(normal)=1.5V ... Unstable: Fails loading windows or reboots when loaded it.



    I'm not pretty sure if I want/must to exceed 1.3625V of Vcore, then I began to search for the same 3.6GHz but with other fsb-mult combination at the same time I want to know what is the higher cpu frecuency I can reach by keeping this Vcore:

    OC: 379x9.5=3.6GHz, Vcore=1.3625V, spd(3.33D)=1263MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(normal)=1.5V ... Unstable: Load windows but fails orthos 1rst test.

    OC: 369x9.5=3.5GHz, Vcore=1.3625V, spd(3.33D)=1230MHz, timmins(manual)=9-9-9-27, CR=2T, Vddr(normal)=1.5V ... Unstable: Load windows but fails orthos 1rst test.

    OC: 358x9.5=3.4GHz, Vcore=1.3625V, spd(3.33D)=1193MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(normal)=1.5V ... OK.- Load windows and PASS orthos 1rst test.



    It seems that as far as I can reach is 3.4GHz with Vcore=1.3625V ... Then I set the cpu at 3.4GHz and test several frequencies for the memories:

    OC: 400x8.5=3.4GHz, Vcore=1.3625V, spd(2.66D)=1066MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(normal)=1.5V ... OK.- Load windows and PASS orthos 1rst test.

    OC: 400x8.5=3.4GHz, Vcore=1.3625V, spd(3.33D)=1333MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(normal)=1.5V ... OK.- Load windows and PASS orthos 1rst test.

    OC: 400x8.5=3.4GHz, Vcore=1.3625V, spd(4.00D)=1600MHz, timmings(manual)=9-9-9-27, CR=2T,Vddr(+0.35V)=1.85V ... Unstable: Load windows but fails orthos 1rst test.



    Finally I tried this settings, as far as I could get within Vcore=1.3625V

    OC: 450x7.5=3.37GHz, Vcore=1.3625V, spd(4.00D)=1800MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(+.40V)=1.95V ... OK.- Load windows and PASS orthos 1rst test.



    Apparently unable to reach 3.6GHz with 1.3625V ... Now I have several questions to overcome the psychological barrier (1GHz + stock):

    Increase the Vcore? Safe limits for 24/7 of it?

    Fix Vcore at 1.3625V and start playing other voltages: FSB (VTT), G-MHC (VCC / NorthBridge), etc? ... I know the function of these voltages but no the utility they have.-

    and better not to speak of more exotic parameters like ... GTL Ref Voltage% ... Clocks Skew Control (CPU / G MHC), etc ...

    WTF ... I know OC is not plug and play, but all the reviews I've read and I have read enough make me suppose this CPU/MB combination should should go further.-

    Greetings.-

  2. #2

    Default Re: E7200 + X48T-DQ6 + DDR3 1800MHz = Only up to 3.4GHz with 1.3625V

    Hi,
    My GA-X48T-DQ6 needs MCH = 1.20V to do cold boot even for small oc, but I'm using Kingston HyperX 1600MHz 7-7-7-2 @1.9V.
    Did you try to add MCH voltage as you up your Vcore.

  3. #3
    Qu4k3r's Avatar
    Qu4k3r is offline Junior Member
    Join Date
    Feb 2009
    Location
    Valencia, VENEZUELA
    Posts
    95

    Default Re: E7200 + X48T-DQ6 + DDR3 1800MHz = Only up to 3.4GHz with 1.3625V

    Quote Originally Posted by Comers View Post
    Hi,
    My GA-X48T-DQ6 needs MCH = 1.20V to do cold boot even for small oc, but I'm using Kingston HyperX 1600MHz 7-7-7-2 @1.9V.
    Did you try to add MCH voltage as you up your Vcore.
    Hi, thank you for answering.

    No, I did not try it. According to next table, values for X48 chipset with 45nm cpu are: MCH volatge: 1.250V(normal) / 1.375V(max), VTT/FSB voltage: 1.10V(normal) / 1.32V(max).

    if I set MHC = 1.20V, then it would be under its normal value, it would decrease instead of rising.

    source: LINK



    ::::::::::::::::::::::::::::::::::::::::::::::::

    Returning to the OC marathon, let me tell you how the situation is. But first I need to redefine the term "stable" that initially it felt like to pass first set/group of orthos tests which last about 16min.

    With my last "stable" settings:
    Quote Originally Posted by Qu4k3r View Post
    OC: 450x7.5=3.37GHz, Vcore=1.3625V, spd(4.00D)=1800MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(+.40V)=1.95V ... OK.- Load windows and PASS orthos 1rst test.
    I test it to see how stable was it and it passed 1rst & 2nd sets of orthos tests which last about 30 min, but orthos fails at 35 min. Now, I must consider "stable" is pass 4 sets of orthos tests at least which last about 1Hrs-5min.

    I don't know what was non-stability source, then I try again keeping all the same but memories:

    OC: 450x7.5=3.37GHz, Vcore=1.3625V, spd(2.66D)=1200MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(normal)=1.5V ... OK.- Load windows and PASS 4 sets of orthos tests. (VTT & MCH both at normal values)



    As well I looked for mininal Vcore value to achieve 3.6GHz without touching neither FSB/VTT voltage nor MHC/NB voltage, but with memory frecuency even lower than previous case in order to avoid instability because it.

    450x8=3.6GHz, spd(2.00D)=800MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(normal)=1.5V

    Vcore -------- Results
    1.36250V --- Fails loading windows or reboots when loaded
    1.36875V --- Fails loading windows or reboots when loaded
    1.37500V --- Fails loading windows or reboots when loaded
    1.38125V --- Fails loading windows or reboots when loaded
    1.38750V --- Load windows but freezes running orthos
    1.39375V --- Load windows but FAILS orthos in less than 30 seg
    1.40000V --- Load windows but FAILS orthos in less than 1 minute


    I don not wanty to set Vcore more than 1.40V, then...
    ┐Must I forget about 3.6GHz?
    Let's suppose that I forgot the 3.6GHz and now I'm just focused on the last really stable configuration:

    OC: 450x7.5=3.37GHz, Vcore=1.3625V, spd(2.66D)=1200MHz, timmings(manual)=9-9-9-27, CR=2T, Vddr(normal)=1.5V

    I want to keep the same cpu frecuency but rise memories to spd(4.00D)=1800MHz 9-9-9-27-2T@1.9V ... and I suppose I have to play with VTT voltage and MHC voltage in
    minimal increases... but, my big question is... ┐How to? ... I mean ...
    What increase in first?
    VTT, MHC, or both at the same time?
    One first and then the other?
    Just move one and leave the other one fixed?

    Here I do not know how the thing goes.-

    MCH allows to increase NB voltage by +0.025V to +0.775 at +0.025V steps.
    MCH (normal) = 1.250v
    MCH (+0.025v) = 1.275v
    MCH (+0.050v) = 1.300v
    MCH (+0.075v) = 1.325v
    MCH (+0.100v) = 1.350v
    MCH (+0.125v) = 1.375v ... Max Value according the table.-

    VTT allows to increase FSB voltage by +0.05v to +0.35V at +0.05V steps.
    VTT (normal) = 1.10V
    VTT (+0.05V) = 1.15V
    VTT (+0.10V) = 1.20V
    VTT (+0.15V) = 1.25V
    VTT (+0.20V) = 1.30V ... Max Value I can set
    VTT (+0.25V) = 1.35V ... NO, becuase exceeds 1.32V recomended by the table

    What MHC/VTT combination must I try? or how must I do the tests?

    Greetings.-
    Last edited by Qu4k3r; 04-09-2009 at 11:50 AM.

  4. #4

    Default Re: E7200 + X48T-DQ6 + DDR3 1800MHz = Only up to 3.4GHz with 1.3625V

    Hi again,
    Sorry, I misspoke.
    My undestanding is that you need to find the minimum MCH votage needed for your ram with moderate oc and keep it unchanged. Apparently my X48T gives MCH normal = 1.00V
    (in contrast to EP45T that gives 1.10V) whick is quite low. In your case I suppose 1.20V should be enough for 2x2GB with up to 2 hard disks.

  5. #5
    Qu4k3r's Avatar
    Qu4k3r is offline Junior Member
    Join Date
    Feb 2009
    Location
    Valencia, VENEZUELA
    Posts
    95

    Default Re: E7200 + X48T-DQ6 + DDR3 1800MHz = Only up to 3.4GHz with 1.3625V

    Quote Originally Posted by Comers View Post
    Hi again,
    Sorry, I misspoke.
    My undestanding is that you need to find the minimum MCH votage needed for your ram with moderate oc and keep it unchanged. Apparently my X48T gives MCH normal = 1.00V
    (in contrast to EP45T that gives 1.10V) whick is quite low. In your case I suppose 1.20V should be enough for 2x2GB with up to 2 hard disks.
    How do you know that? I can't read MHC values at bios in my X48T, I mean I can set it (the increment) only but not read the final value.- Do you have a tool/utility software to read it?
    Last edited by Qu4k3r; 04-08-2009 at 08:52 PM.

  6. #6

    Default Re: E7200 + X48T-DQ6 + DDR3 1800MHz = Only up to 3.4GHz with 1.3625V

    Yes - there is tool: Lavalys EVEREST @ Lavalys - Comprehensive IT Security and Management
    Menu -> Motherboard -> Chipset -> Northbridge -> Core Voltage = ....
    But that 1.00v was only in bios F1 and for all later ver upto F6 normal = 1.05V,
    so to be precise increment +0.20V gives final 1.25V
    Try EVEREST if you didn't so far - you'll love it.

  7. #7
    Qu4k3r's Avatar
    Qu4k3r is offline Junior Member
    Join Date
    Feb 2009
    Location
    Valencia, VENEZUELA
    Posts
    95

    Default Re: E7200 + X48T-DQ6 + DDR3 1800MHz = Only up to 3.4GHz with 1.3625V

    It's stable now!

    I just added FSB VTT & MCH voltages by minimal increment allowed by bios:
    MCH(+0.025v)
    VTT(+0.05v)

    450x7.5=3.375GHz, Vcore=1.3625v, fsb:ram=1:2=spd(4.00D)=1800MHz 9-9-9-27-2T@1.9v

    here the screenshot of 2 hours without errors


    Temps are fine: CPU under 60║C / Mother under 40║C. I'm settled for the moment

    Thanks, greetings.-

  8. #8

    Default Re: E7200 + X48T-DQ6 + DDR3 1800MHz = Only up to 3.4GHz with 1.3625V

    And if you have a cold boot issue just add 1 increment to your MCH voltage only and you'll be solid. Glad you're cruising :)

  9. #9
    Qu4k3r's Avatar
    Qu4k3r is offline Junior Member
    Join Date
    Feb 2009
    Location
    Valencia, VENEZUELA
    Posts
    95

    Default Re: E7200 + X48T-DQ6 + DDR3 1800MHz = Only up to 3.4GHz with 1.3625V

    Quote Originally Posted by Comers View Post
    Yes - there is tool: Lavalys EVEREST @ Lavalys - Comprehensive IT Security and Management
    Menu -> Motherboard -> Chipset -> Northbridge -> Core Voltage = ....
    But that 1.00v was only in bios F1 and for all later ver upto F6 normal = 1.05V,
    so to be precise increment +0.20V gives final 1.25V
    Try EVEREST if you didn't so far - you'll love it.
    But everest shows the same value for MCH, no matter whether it is set to normal or +0.025v


  10. #10

    Default Re: E7200 + X48T-DQ6 + DDR3 1800MHz = Only up to 3.4GHz with 1.3625V

    Yep, indeed Everest shows something different.
    Then Gigabyte's EasyTune should give you the exact reading.

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
  •