Page 102 of 298 FirstFirst ... 2529299100101102103104105112152202 ... LastLast
Results 2,526 to 2,550 of 7441

Thread: ** Official ** Gigabyte UD3 P45 Series -- EP45-UD3 / EP45-UD3R / EP45-UD3P

  1. #2526
    Xtreme Enthusiast
    Join Date
    Oct 2008
    Location
    Florida
    Posts
    562
    Quote Originally Posted by firas View Post
    at this moment my “DRAM Performance Control” is:

    Code:
    Performance Enhance.............................[ Turbo ]
    Extreme Memory Profile (XMP)....................[ Auto  ]
    (G) MCH Frequency Latch.........................[ 333   ]
    System Memory Multiplier........................[ 2.40B ]
    Memory Frequency   ..............1066...... ......1020
    DRAM Timing Selectable..........................[ Manual]
    first it was like this

    Code:
    Performance Enhance.............................[ Turbo ]
    Extreme Memory Profile (XMP)....................[ Auto  ]
    (G) MCH Frequency Latch.........................[ 400   ]
    System Memory Multiplier........................[ 2.0D  ]
    Memory Frequency   ..............1066...... ......890
    DRAM Timing Selectable..........................[ Manual]
    then I tried to follow what PortCitySlim said and what i was in the link I mentioned to higher my ram frequency.

    2.4 multiplier wasn’t available when latch is @ 400, but it was @ latch 333 (and came with the letter B not D)

    i'm thinking of giving 2.5 a shot to bring the frequency a little bit higher than 1066 since my memory voltage now is set to 2.1 (which my shallow knowledge gives me a feeling that my current frequency doesn’t require it )
    At 425fsb you should be using the 266/2.50A. If port sent you my template the only reason its set to 2.00 is because this ram will not run with the 2.40B strap at 482FSB. For me 2.00D at PL9 gives me slighty better everest scores than 2.00B at PL10. Also am I seeing that correctly you have PLL set to 1.15?
    Q9650

    2600k

  2. #2527
    Xtreme Addict
    Join Date
    Jul 2007
    Location
    Arizona - USA
    Posts
    2,200

    Heatpipe logo

    Quote Originally Posted by Dom7184 View Post
    you can take off the NB and heatpipe hold the MOSFET's HS and hold the heatpipe and twist heatpipe it and it should pop lose, i did it with mine and it worked need to take off the v-card to slide out the heat pipe
    It is recommended to uninstall the Heatpipes from the motherboard then use a hair dryer to heat the double sided tape that hold them together and then twist them out.

    Never do this with the heatpipes attached to the mobo.
    //RETIRED-o00o--°(_)°--o00o-OVERCLOCKER//


  3. #2528
    Xtreme Member
    Join Date
    Jan 2009
    Posts
    125
    i've just bought another ud3r, now i've a pair.
    just a question at the moment: how does work xmp? is it really useful?

    @jor3: does domino liquid cooling works well? maybe i'm going to buy it

  4. #2529
    Xtreme Member
    Join Date
    Dec 2007
    Posts
    332
    Quote Originally Posted by Jor3lBR View Post
    It is recommended to uninstall the Heatpipes from the motherboard then use a hair dryer to heat the double sided tape that hold them together and then twist them out.

    Never do this with the heatpipes attached to the mobo.
    well it worked and i didnt feel like taking out the mobo to take off the mofest hs off


    thats why i like my Danger Den Torture Rack

  5. #2530
    Xtreme Member
    Join Date
    Jan 2008
    Location
    Amman, Jordan
    Posts
    140
    Quote Originally Posted by PortCitySlim View Post
    The first thing that catches my eye is your MCH/DRAM Reference. I would adjust this to 1.050 like your channel A/B reference. These settings I gave you last night for 2.1v RAM are exactly the same settings that Hoss331 is using for his overclock stability and they are passing 8+ hour Prime95 blends. I use a lower ram voltage but it lowers down easily enough, for every .02v you lower your RAM you will notice that the reference for that lowers .01v. I am at 2.00v RAM and a 1.000 reference.
    aha ... so the "MCH/DRAM Reference" is somehow connected with A/B reference?
    I’ll set it to 1.050, but do you think I should try to lower the ram voltage a little bit?
    2.060 with 1.030 A/B reference and MCH/DRAM Reference? like what you gave Hoss331 here ?

    Quote Originally Posted by PortCitySlim View Post
    Leaving your CPU TM and No-Execute might not cause you problems(I used to run with them enabled). But remember down the line when you try a blend/OCCT RAM large ftt packet test that disabling these can sometimes solve buggy instability problems.
    ok I’ll disable it, I just thought that those would help me not frying CPU or Memory

    Quote Originally Posted by PortCitySlim View Post
    Setting your performance enhance to turbo might not improve your performance overall, after tweaking you might get better settings to boot on standard. I would say that you should start with Standard and after your finished tweaking everything see if you can run on Turbo
    will set it to standard, I just kept it to Turbo cuz that’s what it was when everything was at stock

    do you think I should set “Static tRead Value” at 8 instead of 9 like your settings? Specially that mine already shows 8 as a default value!

    Quote Originally Posted by Hoss331 View Post
    At 425fsb you should be using the 266/2.50A. If port sent you my template the only reason its set to 2.00 is because this ram will not run with the 2.40B strap at 482FSB. For me 2.00D at PL9 gives me slighty better everest scores than 2.00B at PL10. Also am I seeing that correctly you have PLL set to 1.15?
    didn’t really understand but I’ll give 266/2.50A a shot
    you mean "CPU PLL"? it's @ 1.50 not 1.15

    manny thanks for all the help everyone
    Q9550 E0 @3.8 - zalman CNPS9700 NT - GA-EP45-UD3R rev 1.1 - OCZ Reaper 2x2GB DDR2 1066
    GTX 285 SSC - Thermaltake Toughpower 850W - Samsung 2220WM 22" LCD - win7 x64


  6. #2531
    Xtreme Member
    Join Date
    Oct 2007
    Location
    The Netherlands
    Posts
    110
    Sounds perfect, thank you a lot!
    Silverstone Sugo SG08 | Enermax Modu82+ 625W | ASRock Z77E-ITX | Intel Core i5 3570K /w Indigo Xtreme + CoolIT ECO II ALC + Noctua NF-P14 FLX | 2x4GB Samsung M378B5273DH0-CH9 /w Xilence Heat Pipe Duo | Sapphire HD7870 GE | Kingston HyperX T3 120GB | 2TB Samsung Spinpoint F1

  7. #2532
    Xtreme Mentor
    Join Date
    Nov 2004
    Location
    Elma, NY
    Posts
    3,253
    Finally...had to use 333 strap and GTLs needed to be .838, its prime blend, small and LinX stable over 4.0Ghz.



    i7-2600k @ 4.8Ghz 1.38v L044A892
    ASUS P8P67 LGA
    16GB G.Skill Rip Jaws DDR3 1600Mhz Memory
    1/2" ID Masterkleer, Swiftech MCP-655, Thermochill PA120.3 Rad, XSPC Rasa, MCW-60
    Cooler Master UCP 1100 Watt PSU
    GTX 680 1300/7000
    Lian Li PC V-2000

  8. #2533
    Xtreme Member
    Join Date
    Jan 2009
    Posts
    125
    very well. what about vcore settng in bios? does it vdrop a lot?

  9. #2534
    Xtreme Cruncher
    Join Date
    Jun 2008
    Location
    pacific NW usa
    Posts
    2,764
    if any one is interested im selling my two g.skill pc8800 kits 2x2gb for a total of 8 gigs and my E8500 EO Q827A272 in the F/S section.i just got my first set of g.skill 9600's and soon to be getting another,so no need for the 8800's.
    _________________________________________________
    ............................ImAcOmPuTeRsPoNgE............................
    [SIGPIC][/SIGPIC]

    MY HEATWARE 76-0-0

  10. #2535
    Xtreme Member
    Join Date
    Jan 2008
    Location
    Michigan
    Posts
    313
    Quote Originally Posted by dnottis View Post
    It came with F6 but so far F8F has been better to me. If I run into issues I'll try F4. What were you able to do with F4 that was giving you problems with the newer bioses? Higher Fsbs? Stability?
    Better stability and higher FSB. As a reference test, I used F4 and kept everything on auto except vcore, PLL, and VTT. I was able to run 471x8.5 P95 stable for an hour. Every other BIOS with the exact same settings (vcore, PLL, and VTT were manually set to the same options as I did in F4) failed P95 within 10 mins.

  11. #2536
    Xtreme Mentor
    Join Date
    Nov 2004
    Location
    Elma, NY
    Posts
    3,253
    Quote Originally Posted by ghost_recon88 View Post
    Better stability and higher FSB. As a reference test, I used F4 and kept everything on auto except vcore, PLL, and VTT. I was able to run 471x8.5 P95 stable for an hour. Every other BIOS with the exact same settings (vcore, PLL, and VTT were manually set to the same options as I did in F4) failed P95 within 10 mins.
    I switched to F4 to hit what I did above. Thanks for the tip seems to be working well for me!

    i7-2600k @ 4.8Ghz 1.38v L044A892
    ASUS P8P67 LGA
    16GB G.Skill Rip Jaws DDR3 1600Mhz Memory
    1/2" ID Masterkleer, Swiftech MCP-655, Thermochill PA120.3 Rad, XSPC Rasa, MCW-60
    Cooler Master UCP 1100 Watt PSU
    GTX 680 1300/7000
    Lian Li PC V-2000

  12. #2537
    Xtreme Member
    Join Date
    Dec 2007
    Posts
    332
    Quote Originally Posted by dnottis View Post
    I switched to F4 to hit what I did above. Thanks for the tip seems to be working well for me!
    hmm... i think i wanna try it out also mine cam with the F6 for my Q

    cuz does it flash back easy or i need to do the same crap with asus mobo to flash to an older bios ?

    also anyone know which one os good for dual cores [E8500] ?

  13. #2538
    Xtreme Mentor
    Join Date
    Nov 2004
    Location
    Elma, NY
    Posts
    3,253
    Quote Originally Posted by Dom7184 View Post
    hmm... i think i wanna try it out also mine cam with the F6 for my Q

    cuz does it flash back easy or i need to do the same crap with asus mobo to flash to an older bios ?

    also anyone know which one os good for dual cores [E8500] ?
    Put F4 on a flash drive, reboot, go into bios hit F8 for Qflash and flash it like using AltF2 on an Asus, was cake and no issues going back either.

    i7-2600k @ 4.8Ghz 1.38v L044A892
    ASUS P8P67 LGA
    16GB G.Skill Rip Jaws DDR3 1600Mhz Memory
    1/2" ID Masterkleer, Swiftech MCP-655, Thermochill PA120.3 Rad, XSPC Rasa, MCW-60
    Cooler Master UCP 1100 Watt PSU
    GTX 680 1300/7000
    Lian Li PC V-2000

  14. #2539
    Xtreme Member
    Join Date
    Dec 2007
    Posts
    332
    Quote Originally Posted by dnottis View Post
    Put F4 on a flash drive, reboot, go into bios hit F8 for Qflash and flash it like using AltF2 on an Asus, was cake and no issues going back either.
    thanks right now im on F7 but wanted to try F4 since it should be better for Q BRB

  15. #2540
    Xtreme Member
    Join Date
    Jan 2009
    Posts
    125
    Quote Originally Posted by ghost_recon88 View Post
    Better stability and higher FSB. As a reference test, I used F4 and kept everything on auto except vcore, PLL, and VTT. I was able to run 471x8.5 P95 stable for an hour. Every other BIOS with the exact same settings (vcore, PLL, and VTT were manually set to the same options as I did in F4) failed P95 within 10 mins.
    i've a xeon x3370 under bios F5. i should try to flash F4 for some tests

  16. #2541
    Xtreme Member
    Join Date
    Dec 2007
    Posts
    332
    @dnottis what are your settings ?

    also can you try 8X500 ?

    Load-Line Calibration
    CPU Vcore
    CPU Termination
    CPU PLL
    CPU Reference

    MCH Core
    MCH Reference
    MCH/DRAM Refernce
    ICH I/O
    ICH Core

    DRAM Voltage
    DRAM Termination
    Channel A Ref
    Channel B Ref

  17. #2542
    Registered User
    Join Date
    Jan 2007
    Posts
    67
    I also am currently trying F4

    So far its good
    Last edited by SpeedEuphoria; 03-04-2009 at 12:23 PM.
    Q9650
    Gigabyte UD3P, OCZ 2x2GB 1066
    Asus Top 4850 + Accelero S1 rev2
    2x WD Raptor 74gb raid0 + 1TB storage
    Antec TP3 Trio 550w, Windows XP
    Aerocool Spiral Galaxies Modded
    37"Olevia 537H

  18. #2543
    Registered User
    Join Date
    Jan 2009
    Posts
    61
    Quote Originally Posted by PortCitySlim View Post
    If you benchmark under disabled and then under enabled and compare that to your results on auto you should be able to detect what the auto setting is doing.
    I've just benchmarked with Everest and there's absolutely no difference between enabled/disabled/auto.

  19. #2544
    Xtreme Member
    Join Date
    Jan 2009
    Posts
    125
    Quote Originally Posted by Touche View Post
    I've just benchmarked with Everest and there's absolutely no difference between enabled/disabled/auto.
    i've tried too. no difference.

  20. #2545
    Xtreme Mentor
    Join Date
    Nov 2004
    Location
    Elma, NY
    Posts
    3,253
    Load-Line Calibration - enabled
    CPU Vcore - 1.34xx (dont recall which one, but it's one of the 1.34 options)
    CPU Termination - 1.26v
    CPU PLL - 1.57
    CPU Reference - .838

    MCH Core 1.26v or 1.28v
    MCH Reference .838
    MCH/DRAM Refernce AUTO
    ICH I/O - lowest selectable
    ICH Core - lowest selectable

    DRAM Voltage - 1.82v
    DRAM Termination - AUTO
    Channel A Ref - AUTO
    Channel B Ref - AUTO

    also can you try 8X500 ?
    Yep but my chip doesn't like 500. It boots but it's very unstable.

    i7-2600k @ 4.8Ghz 1.38v L044A892
    ASUS P8P67 LGA
    16GB G.Skill Rip Jaws DDR3 1600Mhz Memory
    1/2" ID Masterkleer, Swiftech MCP-655, Thermochill PA120.3 Rad, XSPC Rasa, MCW-60
    Cooler Master UCP 1100 Watt PSU
    GTX 680 1300/7000
    Lian Li PC V-2000

  21. #2546
    Xtreme Member
    Join Date
    Jan 2009
    Posts
    125
    i suggest to manually set the mch gtl and to increase cpu gtl

  22. #2547
    Xtreme Member
    Join Date
    Jul 2008
    Location
    Louisiana
    Posts
    151
    Quote Originally Posted by firas View Post
    aha ... so the "MCH/DRAM Reference" is somehow connected with A/B reference?
    I’ll set it to 1.050, but do you think I should try to lower the ram voltage a little bit?
    2.060 with 1.030 A/B reference and MCH/DRAM Reference? like what you gave Hoss331 ?
    Sure lower it away, if it passes keep lowering it



    Quote Originally Posted by firas View Post
    will set it to standard, I just kept it to Turbo cuz that’s what it was when everything was at stock

    do you think I should set “Static tRead Value” at 8 instead of 9 like your settings? Specially that mine already shows 8 as a default value!
    SURE! That is what I was meaning about using standard and trying better Tread values and write training. Set your to stardard and see how low your Tread value will boot, that is after you get your multi and ram speed set to your liking.



    Quote Originally Posted by firas View Post
    didn’t really understand but I’ll give 266/2.50A a shot
    you mean "CPU PLL"? it's @ 1.50 not 1.15

    manny thanks for all the help everyone
    I would follow Hoss331's advice with setting a higher RAM Multi, Hoss helps me with my settings and is pretty good at this stuff.
    Q9650@4.25ghz 500x8.5 1.37v
    Gigabyte GA-EP45-UD3P - F9B
    Gskill PI PC2-8800 4gb 1000mhz 55ns 1.92v
    WD Caviar Black 640gb/Kingston HyperX 8gb 0.7ms Access Storage
    Patriot XporterXt 32gb 0.8ms Access Storage
    Sapphire HD 3870 512mb
    Zalman 750watt PSU
    NZXT Tempest with 5x120mm+2x140mm
    Sunbeam "Core Contact"/Scythe "Slipstream" 110cfm/
    IC Diamond

  23. #2548
    Xtreme Member
    Join Date
    Jan 2008
    Location
    Michigan
    Posts
    313
    Quote Originally Posted by dnottis View Post
    Load-Line Calibration - enabled
    CPU Vcore - 1.34xx (dont recall which one, but it's one of the 1.34 options)
    CPU Termination - 1.26v
    CPU PLL - 1.57
    CPU Reference - .838

    MCH Core 1.26v or 1.28v
    MCH Reference .838
    MCH/DRAM Refernce AUTO
    ICH I/O - lowest selectable
    ICH Core - lowest selectable

    DRAM Voltage - 1.82v
    DRAM Termination - AUTO
    Channel A Ref - AUTO
    Channel B Ref - AUTO

    also can you try 8X500 ?
    Yep but my chip doesn't like 500. It boots but it's very unstable.
    Use .63x as the multipler for figuring out GTLs.

    VTTx.63=GTL voltage

    So, if VTT is 1.20, do 1.20x.63=75.6v, or in this case you would chose .76v in the GTL list. Same method and format goes for the NB. Using that formula I was able to get a CPU-Z validation with my Q9550 at 555FSB on air

  24. #2549
    Xtreme Member
    Join Date
    Jul 2008
    Location
    Louisiana
    Posts
    151
    Quote Originally Posted by firas View Post
    ok I’ll disable it, I just thought that those would help me not frying CPU or Memory

    I would not worry about the Thermal Monitor and your current overclock. Your not going to experience high load temps so Thermal Monitor is going to be useless for you. Even at my clock speeds sometimes I do get my chip a bit hot(If I linpack) but the important thing to do is monitor your temps and even if they get above 70C as long as your not running that hot for a long period of time your not going to "fry" one of these Q9550/Q9650 chips.
    Q9650@4.25ghz 500x8.5 1.37v
    Gigabyte GA-EP45-UD3P - F9B
    Gskill PI PC2-8800 4gb 1000mhz 55ns 1.92v
    WD Caviar Black 640gb/Kingston HyperX 8gb 0.7ms Access Storage
    Patriot XporterXt 32gb 0.8ms Access Storage
    Sapphire HD 3870 512mb
    Zalman 750watt PSU
    NZXT Tempest with 5x120mm+2x140mm
    Sunbeam "Core Contact"/Scythe "Slipstream" 110cfm/
    IC Diamond

  25. #2550
    Registered User
    Join Date
    Jan 2009
    Posts
    61
    Quote Originally Posted by simone18 View Post
    i've tried too. no difference.
    Done some more testing - no difference in Everest benchmark but there is in stability. DDR Write Training Enabled gives me Linpack errors. I've set it to disabled.

Page 102 of 298 FirstFirst ... 2529299100101102103104105112152202 ... LastLast

Tags for this Thread

Bookmarks

Bookmarks

Posting Permissions

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