Page 1 of 3 123 LastLast
Results 1 to 25 of 74

Thread: GIGABYTE X58A-OC First Touch and Preliminary Performance Review

  1. #1
    PIfection
    Join Date
    May 2006
    Posts
    989

    GIGABYTE X58A-OC First Touch and Preliminary Performance Review

    Introduction

    ASUS have for a long time been producing their ROG series that we all feel is very well targeted to overclockers. Recently Gigabyte got into the game with their release of the high end X58A-UD9 board, but where both of these platforms fall short is being designed from the ground up by overclockers for overclockers, and this is where the X58A-OC steps in.

    From the pictures already released you can see this board means business, with an IO panel that looks like it lost a leg in the war to features such as OC-VRM and OC-Touch it's clear this board has potential to be a serious contender.

    I don't have the extreme performance I know your all hungry for quite yet but today is just a quick look at performance and a demonstration of the OC Touch feature set. Later I will investigate the extreme performance from this board when paired with an Intel 990X processor.

    I have picked a few benchmarks that I like to use when comparing performance on different platforms. Vantage Performance, SuperPI 8m, 3DMark03 and Everest. These four give a good overview of motherboard performance and any areas a board might excel or have problems.



    Test Platform
    GIGABYTE X58A-OC @ 4GHz, 3276 MHz uncore
    Intel 980X Gulftown
    GSKILL Perfect Storm 2000 9-9-9-24 @ 910 8-9-8-24
    CORSAIR Force SSD 60GB
    GIGABYTE HD 5870

    All benchmarks are run on the same clean operating system, the same versions of software and drivers and with no addition tweaking done in BIOS.



    Preliminary Performance Analysis

    Vantage Performance



    A strong note to start on, despite what looks to be a big lead on the graph, there is not a lot in it, but the OC does take the lead. This is an important result as the X58A-UD9 was a little of the mark when it comes to 3DMark scores, a little behind ASUS. Gigabyte hits back hard here with efficiency that is capable of taking world records



    SuperPI 8M



    So with the SuperPI 8M graph the lower the seconds taken to complete 8M, the better the result. Again here the results are in favor of the OC, followed closely behind by the Sapphire PURE X58 which I have mentioned is very efficient out of the box in the past.

    The X58A-UD9 which I will continue comparing back to throughout this report is an extremely fast board for SuperPI and while the OC is a "TOUCH" behind when I compared them directly it is only by under 100ms over 3 separate runs. I would be confident this board could take out low clock challenges, I guess we will shortly find out!



    3DMark03



    In 3DMark03 the Sapphire board pulls a nice figure that puts it into first place but the OC is only 30 marks behind. Again this is an encouraging result as the UD9 was a few hundred marks behind when it came to single card 3DMark03.



    Everest



    There is no shame losing to an ASUS board in Everest bandwidth as ASUS bios are always apply tight settings from the out set. This difference of 20 MB/s between all 3 boards shows there is nothing in it with all boards on a level playing field for write bandwidth.

    With this result in mind it does explain the 200 milliseconds the Sapphire is behind the Gigabyte in 32m, but doesn't go anywhere to explaining the R3E being 4 seconds off the pace. As can be seen on low clock challenge threads when the R3E is tightened up it is arguably the fastest X58 board on the planet... but for how long?



    OC-Touch Demonstration

    This feature kicks serious ass, Gigabyte have been slow on the uptake with a hardware style frequency control but I love the multiplier adjustment, this will change the way I bench. No more having to worry about software hotkeys for multiplier adjustments. The keyboard hooks running in the background polling the system every however many milliseconds for keystrokes the better.





    Results

    Here are the screen shots from the above results plus a few more I ran

    Vantage X looks good , a little more efficient than the Sapphire X58, where it should be for a board aiming to be the premier overclocking platform in the world




    some high clock 32M with the Corsair GTX6 sticks and CPU on air



  2. #2
    I am Xtreme
    Join Date
    Aug 2008
    Posts
    5,586
    was looking at the last two screenshots...the amount of ram not readable..is it 2048mb/single channel?


  3. #3
    PIfection
    Join Date
    May 2006
    Posts
    989
    its two 1GB modules dude lol, corsair gtx6 is 1GB module hehe

  4. #4
    I am Xtreme
    Join Date
    Dec 2004
    Location
    France
    Posts
    3,462
    nice job man! It looks to be a really nice board, i can't wait to get mine.

    you did disable the memory executable thing in the bios right? for superpi?

    BTW it is very nice to see this board own in 2D and even in 3D that is impressive.
    Last edited by sin0822; 02-23-2011 at 01:22 PM.

  5. #5
    I am Xtreme
    Join Date
    Aug 2008
    Posts
    5,586
    oh gtx6 i thought it was gtx2

    why not tripple channel?


  6. #6
    PIfection
    Join Date
    May 2006
    Posts
    989
    well just to clarify all the test results except that SuperPI 32M are done with GSKILL Perfect Storm 2000 kit, the 32m has been done with GTX6 2x1GB modules for a few reasons,

    firstly i only have 2 gtx6 sticks,

    but also i think they are the way to go for 990X, being 1GB modules, they are much lighter on the IMC,

    so when you are looking at clocking your 990X to 6.6 on LN2 you will probably need single channel, as you would have noticed from all the results out so far, but maybe with a kit of 2x 1GB modules that are easier on the IMC you can use 2GB of ram and for benchmarks like AM3, PIFAST even 3D05 this is important!

    real life scenario testing you could say..

  7. #7
    PIfection
    Join Date
    May 2006
    Posts
    989
    Quote Originally Posted by sin0822 View Post
    you did disable the memory executable thing in the bios right? for superpi?
    hey mate,

    nope my prelim performance testing is always done without touching anything in bios except making the 3 platforms frequency, uncore, qpi and memory frequency/timings identical... all other settings are left as default (except of course disabling some of the unrequired ethernet ports, c1, etc)

    i want to get an idea of performance right out of the box to start with, then move on to tightening things up from there

  8. #8
    I am Xtreme
    Join Date
    Aug 2008
    Posts
    5,586
    ok so you only have 2 sticks...i know we need 3 sticks for pifection32


  9. #9
    Aussie God
    Join Date
    Feb 2005
    Location
    Copenhagen, Denmark
    Posts
    4,596
    Hey James,
    Can you please tell a bit more about accurracy of your test..
    Results are really close, could be random...

    What did you do? Did you take average of X runs on each product?
    Hope you dont mind me asking, just like this clarified when it is very close results.

    Also can you please post CPU-Tweaker, or other verification, to let us know aggressivity of ram subbies per default versus the other products.

    Cheers,
    Marc
    Competition ranking;
    2005; Netbyte, Karise/Denmark #1 @ PiFast
    2008; AOCM II, Minfeld/Germany #2 @ 01SE/AM3/8M (w. Oliver)
    2009; AMD-OC, Viborg/Denmark #2 @ max freq Gigabyte TweaKING, Paris/France #4 @ 32M/01SE (w. Vanovich)
    2010: Gigabyte P55, Hamburg/Germany #6 @ wprime 1024/SPI 1M (w. THC) AOCM III, Minfeld/Germany #6 @ 01SE/AM3/1M/8M (w. NeoForce)

    Spectating;
    2010; GOOC 2010 Many thanks to Gigabyte!


  10. #10
    I am Xtreme
    Join Date
    Dec 2004
    Location
    France
    Posts
    3,462
    Quote Originally Posted by youngpro View Post
    hey mate,

    nope my prelim performance testing is always done without touching anything in bios except making the 3 platforms frequency, uncore, qpi and memory frequency/timings identical... all other settings are left as default (except of course disabling some of the unrequired ethernet ports, c1, etc)

    i want to get an idea of performance right out of the box to start with, then move on to tightening things up from there
    gotcha, that makes sense

    And yea single channel for subzero to get around cold bug too.

  11. #11
    Xtreme Legend
    Join Date
    Mar 2005
    Location
    Australia
    Posts
    17,242
    LOL @ OC-Peg
    Team.AU
    Got tube?
    GIGABYTE Australia
    Need a GIGABYTE bios or support?



  12. #12
    Xtreme Addict
    Join Date
    Nov 2002
    Location
    Houston
    Posts
    1,123
    Quote Originally Posted by youngpro View Post
    hey mate,

    nope my prelim performance testing is always done without touching anything in bios except making the 3 platforms frequency, uncore, qpi and memory frequency/timings identical... all other settings are left as default (except of course disabling some of the unrequired ethernet ports, c1, etc)

    i want to get an idea of performance right out of the box to start with, then move on to tightening things up from there
    So you enabled Microcode ignore option on the Sapphire and OC boards but not on the RIIIE, but then disabled it for the 3D runs on those same boards? That option is what allows you to drop the SuperPI times down to those results but then 3D will suffer if it is left enabled.

    I assume you left PM settings alone as the setups between the boards (assume OC is similar to the UD9 in regards to BIOS setup). If those options are not the same between each board, then the tests are not truly apples to apples.
    Why do people claim to know something they do not and then write about it.

  13. #13
    Xtreme X.I.P.
    Join Date
    Nov 2005
    Location
    Sweden
    Posts
    671
    Looks like a solid board. But seriously James, those scales are retarded Difference of 0.08% and it shows as twice the size.
    2008 - AOCC WW #2 | MOA EU #8 | GOOC WW #1
    2009 - GOOC WW #3 | MOA EU #3 | MOA WW #1 | GB TweaKing #6 | ASUS ROG OCS #2
    2010 - MOA EU #1 | GOOC EU #13 | MOA WW #1
    2011 - MOA EU #4 | MOA WW #?

  14. #14
    Xtreme Legend
    Join Date
    Mar 2005
    Location
    Australia
    Posts
    17,242
    Gary microcode ignore only improves efficiency on A0 Gulftown, not on B1
    Quote Originally Posted by bingo13 View Post
    So you enabled Microcode ignore option on the Sapphire and OC boards but not on the RIIIE, but then disabled it for the 3D runs on those same boards? That option is what allows you to drop the SuperPI times down to those results but then 3D will suffer if it is left enabled.

    I assume you left PM settings alone as the setups between the boards (assume OC is similar to the UD9 in regards to BIOS setup). If those options are not the same between each board, then the tests are not truly apples to apples.
    Last edited by dinos22; 02-23-2011 at 03:24 PM.
    Team.AU
    Got tube?
    GIGABYTE Australia
    Need a GIGABYTE bios or support?



  15. #15
    I am Xtreme
    Join Date
    Dec 2004
    Location
    France
    Posts
    3,462
    LOL i just watched that video, the ending was the best. LOL

    hey gary.

  16. #16
    PIfection
    Join Date
    May 2006
    Posts
    989
    Quote Originally Posted by bingo13 View Post
    So you enabled Microcode ignore option on the Sapphire and OC boards but not on the RIIIE, but then disabled it for the 3D runs on those same boards? That option is what allows you to drop the SuperPI times down to those results but then 3D will suffer if it is left enabled.
    er no i didnt touch the microcode options at all, and fyi even if i did touch any microcode options, it only works for a0, there is no performance gains for b1 chips, you really think i am doing bias review? if i am doing bias review, why would i post 1 graph where r3e wins and 1 graph where sapphire x58 wins... it doesnt make sense?

    i dont mind you having a go at me if i do something wrong mate, but at least get your facts straight before you do it

    and yes Jon I know the scales are retarded, i am sux at excel
    Last edited by youngpro; 02-23-2011 at 03:13 PM.

  17. #17
    PIfection
    Join Date
    May 2006
    Posts
    989
    this is just average over 2 runs Marc, yes the close results could average out even closer over time for sure, the Sapphire could jump in front in Vantage or 8m, but also the X58A-OC could jump in front in Everest or 03 also..

  18. #18
    Registered User
    Join Date
    Dec 2002
    Location
    Somewhere in Sth East Asia
    Posts
    170
    Thanks for the write up James. Even if your excel sux

    Nice features on that board - will definitely be giving one of these a spin. Love that on the fly multi adjustment

    Quote Originally Posted by sin0822 View Post
    LOL i just watched that video, the ending was the best. LOL.
    heh not my idea of a "happy ending" Still scares the living bejesus out of me

  19. #19
    Xtreme Legend
    Join Date
    Mar 2005
    Location
    Australia
    Posts
    17,242
    hey James is it going to be cold in Adelaide next month for some mankini action lmao nvm imagine watching a V8 supercars race wearing a mankini, LOL
    Team.AU
    Got tube?
    GIGABYTE Australia
    Need a GIGABYTE bios or support?



  20. #20
    Registered User
    Join Date
    Dec 2002
    Location
    Somewhere in Sth East Asia
    Posts
    170
    Quote Originally Posted by dinos22 View Post
    hey James is it going to be cold in Adelaide next month for some mankini action lmao nvm imagine watching a V8 supercars race wearing a mankini, LOL
    I'll pay $$$$$$ to see you walking around the general admin area like that. We can run a sweepstake to see how long you last

  21. #21
    PIfection
    Join Date
    May 2006
    Posts
    989
    lol for sure but the money we collect will need to go into paying the gap for dinos private health insurance after he gets the crap beaten out of him

  22. #22
    Xtreme Legend
    Join Date
    Mar 2005
    Location
    Australia
    Posts
    17,242
    i can already see the enraged bogans, some red faces in anger and others puking at the site of mankini
    Team.AU
    Got tube?
    GIGABYTE Australia
    Need a GIGABYTE bios or support?



  23. #23
    I am Xtreme
    Join Date
    Dec 2004
    Location
    France
    Posts
    3,462
    Quote Originally Posted by youngpro View Post
    lol for sure but the money we collect will need to go into paying the gap for dinos private health insurance after he gets the crap beaten out of him
    LOL make sure to pay a nice deductible too lol.

    pro- excel sucks it automatically scales like this when you downsize the graph.

    Anyways I can see this board ruling the 2D scene like no other.

  24. #24
    I am Xtreme
    Join Date
    Dec 2008
    Location
    France
    Posts
    9,060
    What the f*** is wrong with you, people?! So I open the thread to read a review and see this...
    Quote Originally Posted by elmor View Post
    Looks like a solid board. But seriously James, those scales are retarded Difference of 0.08% and it shows as twice the size.
    What do you know? It's 1337 PR and marketing skillz!
    Gotta say the SPi difference is significant, though.
    Last edited by zalbard; 02-23-2011 at 04:31 PM.
    Donate to XS forums
    Quote Originally Posted by jayhall0315 View Post
    If you are really extreme, you never let informed facts or the scientific method hold you back from your journey to the wrong answer.

  25. #25
    Xtreme Member
    Join Date
    Jul 2009
    Location
    Brisbane, Australia
    Posts
    182
    Any word on the release date?

Page 1 of 3 123 LastLast

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
  •