[H] 680i & Yorkfield Support Statement from NV

Discussion in 'Motherboards' started by FrgMstr, Nov 28, 2007.

  1. sd11

    sd11 [H]ard|Gawd

    Messages:
    1,141
    Joined:
    Apr 11, 2007
    Yeah, which begs the question. Since the 780i is a 680i with a pci-e 2.0 slapped on and out of the box tri-SLI... is it really an electrical issue with the 680i, is the chipset simply unable to do it period, or did they screw up the 780i as well.:rolleyes:
     
  2. drpeppir

    drpeppir Limp Gawd

    Messages:
    165
    Joined:
    Mar 21, 2004
    Well the 780i isn't rlsed yet, so they have time for Yorkie compatibility revisions im sure, but the differences between the revised board and the originally planned product may result in stability issues again like the 680i due to inadequate testing.
     
  3. Dan_D

    Dan_D [H]ard as it Gets

    Messages:
    53,424
    Joined:
    Feb 9, 2002
    The truth is we don't know what's wrong with all the 680i SLI boards. Obviously the explanation that makes the most sense relates to inadequate testing, and certainly the 680i SLI chipset based boards all feel like rushed designs.

    We can only hope that the 780i SLI is better.
     
  4. BillParrish

    BillParrish [H]ardness Supreme

    Messages:
    7,520
    Joined:
    Aug 25, 2006
    I am at a loss to discover where I said it was straight forward. Perhaps I should have put a :rolleyes: in. My bad. Actually it is pretty straight forward to make board mods if you know what you are doing and have the skill, but read my post again. As you pointed out, that was not really the point. We are not talking about someone fooling around not knowing what they are doing, a la XS. I am talking about mods where the engineers have circuit schematics available and board internal layout information/AW and can determine were exactly to "pick off" connection points to modify a circuit in an exact manner, test the results and then create procedures that a skilled factory worker (read a woman who has been trained in proper soldering and repair techniques) can implement the modification easily. That is how it is done in the real world, I was one of the engineers. You cannot trash finished product because of an "oh shit", its not so much the cost, its the customer delivery schedule. But that was telecom.

    I guessed surprised was too weak of a word on my part. I highly suspect that there is a fundamental design flaw not easily fixable. That is why they just do not announce "oops sorry, here is a rev 2.0 board that will work." While that would not help current owners it does indicate they have a handle on the issue and are moving forward. That appears not to be the case.

    However I would love to know from an engineering perspective exactly what is wrong even if they do not know or cannot find a solution for current boards. Shit happens, been there got the scars. I highly suspect the VRM as you allude. What is interesting is that the quads and the duals share the same datasheet which having read it I can find no difference in the quad other than its power requirements. However I may have missed something and from what I gather there is proprietary info Intel does not make public so /shrug.

    http://www.intel.com/design/processor/datashts/313278.htm
     
  5. Random Guy

    Random Guy [H]Lite

    Messages:
    69
    Joined:
    Jul 12, 2007
    Has this issue been resolved yet? I have received an e-mail from XFX Support regarding my 680i LT mobo:-

    ...we do expect a BIOS update to be available soon to support the Yorkfield CPU`s, with the exception of the Yorkfield Quadcore Extreme. Regards...
     
  6. MrWizard6600

    MrWizard6600 [H]ardness Supreme

    Messages:
    5,777
    Joined:
    Jan 15, 2006
    How has your experiance been with 680i. I understand you've had to Rma three (3) strikers. Thats pretty horribad. And then you've got your hands on an evga black pearl and stock evga 680i board. How have they been?
     
  7. Dan_D

    Dan_D [H]ard as it Gets

    Messages:
    53,424
    Joined:
    Feb 9, 2002
    Actually I've owned 11 EVGA 680i SLI boards and only one Striker Extreme. I have also worked with the P5N32-E SLI and additional EVGA 680i SLI boards that were used in the construction of various machines for friends of mine. I also reviewed the Gigabyte 680i SLI board.

    The first four EVGA 680i SLI boards I owned were all the initial boards that were produced before BIOS P23 and P24 were released. They had problems with the Sound Blaster Creative X-Fi as well as problems with the onboard sound. One of them turned out to be DOA. I ditched those before the sound issues were solved. I kept the Striker Extreme used in our review and that was in my machine for some time. (It too had sound issues that ASUS didn't solve until three our four months after the release. I used a cheap Sound Blaster Live! card that fixed my issues as it too had issues with the X-Fi. After awhile I tried a memory upgrade and my at the time very expensive modules wouldn't work at all with the Striker. So after being disappointed in the overclocking capabilities and the fact that I couldn't use my expensive memory modules I went ahead and switched to another EVGA 680i SLI board. Originally I had no issues with it and it worked for a couple months. Fast foward a bit and I purchased a Q6600 and found out first hand about the quad core overclocking issues. So I decided to not wait for an RMA and I used that AR board in my GF's machine and I bought a brand spanking new A1 board for my own rig. So that means I was already on board number 6 in about as many months. Though only my 5th EVGA board. After just under a month the board burned up and quit working. So I was still within my return policy at Fry's so I simply swapped the board out. Next board was killed by a leak in my water cooling setup. That's obvioulsy not the fault of the board's design. Next board was purchased and then broken accidentally because I dropped a screwdriver directly onto the board and broke a capacitor off the board.

    I decided that since I was water cooling I'd pull the board and test it later. If it worked I'd sell it and I would simply use the next one in my build. So I purchased another board and that board was DOA. So I took it back. Then I got yet another board which burned out in under a month. So I returned that one and the next EVGA board burned up in two weeks. So I swapped for yet another board which worked for the last two months then died. So now I'm on number 11 which is the EVGA Black Pearl. It's working great so far, but I never underestimate my ability to kill these things. The one board that I broke the capacitor off I sold for cheap to a friend and it still works today, though it too is showing signs of impending failure. NB temps shot up to 80c on average and it is experiencing USB issues. Below are my boards in a list format and what happened to them.

    List of boards I've owned:

    1.) 122-CK-NF68-AR Crackling audio problem. (Later fixed in BIOS P23&P24) -Returned to retailer.
    2.) 122-CK-NF68-AR Crackling audio problem, stability issues. -Returned to retailer.
    3.) 122-CK-NF68-AR DOA. -Returned to retailer.
    4.) 122-CK-NF68-AR Crackling audio problem, stability issues, SATA corruption problems. -Returned to retailer for refund.
    5.) 122-CK-NF68-AR Purchased after months of using the Striker Extreme. Used with great success and works today. -Replaced with 122-CK-NF68-A1 for Q6600 overclocking.
    6.) 122-CK-NF68-A1 Purchased to replace previous AR board for Q6600 overclocking. Board fried after less than one month of use. -Returned to retailer.
    7.) 122-CK-NF68-A1 Board killed by leak in cooling system. -Destroyed.
    8.) 122-CK-NF68-A1 Board damaged by accident. Board was removed from system and stored. (Board later tested ok and sold to a friend.)
    9.) 122-CK-NF68-A1 Board died within three weeks of usage. -Returned to retailer.
    10.) 122-CK-NF68-A1 Board worked for two months and then died. -Scrapped
    11.) 122-CK-NF69-A1 EVGA Black Peal Edition currently in use.

    So basically I've had a number of these things for several reasons. All the earliest ones were purchased and returned before BIOS P23 and P24 came out so I don't know if those boards were truly bad or not. Out of the three EVGA 680i SLI boards I have currently only two of them work. My Striker Extreme still works but I don't use it due to problems with memory compatibility and it is an extremely poor overclocker. It is ok with dual core processors and horrid with quads. My experiences with the Gigabyte N680SLI-DQ6 was a good board but I never had the desire to purchase one. As for the P5N32-E SLI it is just as bad as the Striker Extreme. BIOS flashing it is still as scary as flashing the Striker and I'm sick of replacing BIOS chips on those things due to either that crappy Windows utility or the usual causes of flashing failures. I've never in my life seen boards that have so many issues accepting BIOS flashes.

    So that's been my 680i SLI experience over the last year. :eek:
     
  8. Jodiuh

    Jodiuh [H]ardness Supreme

    Messages:
    5,275
    Joined:
    Feb 3, 2004
    I have the A! revision. It squeals, whines, and screams when stressing ram...or scrolling the mouse wheel. :p
     
  9. Dan_D

    Dan_D [H]ard as it Gets

    Messages:
    53,424
    Joined:
    Feb 9, 2002
    That sucks. :confused:
     
  10. silent-circuit

    silent-circuit [H]ardForum Junkie

    Messages:
    16,133
    Joined:
    Sep 18, 2005
    They'll let you RMA for that, but it just seems like an awful lot of work to me. Then again, for most I doubt "drain the water loop (curse)" is step 1 in removing the motherboard from their case.
     
  11. Dan_D

    Dan_D [H]ard as it Gets

    Messages:
    53,424
    Joined:
    Feb 9, 2002
    I'm almost sorry I went to water cooling simply because taking this rig apart so many times has been such a huge pain in the ass.
     
  12. STEM

    STEM Gawd

    Messages:
    561
    Joined:
    Jun 7, 2007
    Before I trow in my two cents, I would like to say to Dan_D that doing the same thing over and over again and expecting different results is the definition of insanity. But I do understand your reasons, you wanted SLI support. Second of all, while the Striker Extreme provides piss poor overclocking for a quad core, allong with poor support for RAM, the P5N32-E SLI is actually a better board. ASUS constantly updates the BIOS on it, and I can get 3.4 ~ 3.6 Ghz with a Q6600 rev. G0. The Striker's BIOS hasn't been updated in ages, and with BIOS 1305 I couldn't get more than 2.8 out of a Q6600. In regards to the EVGA 680i and all the reference design boards manufactured by FOXCONN for NVidia, the reason why they fail so badly is because of the poor build quality. Cheap caps, cheap components, allong with piss poor assembly is the reason why those boards fails so easily. And I don't know if you guys where aware that allot of the older EVGA 680i boards contain LED. I've found that on EVGA's web site, and they recommend to wash your hands and what not after working with one of those boards. Anyway. couple the poor construction of those boards with the high power demand that an overclocked Quad Q6600 (even rev. G0) is putting on it, and you'll see why they fail. :eek:

    And now to the issue at hand:
    I've read every possible conspiracy theory out there regarding Penryn support on 680i boards. Yet no one has taken into consideration that maybe it's not Intel, but NVidia that doesn't want Yorksfield to be supported with their current chipsets. I don't know if anybody else noticed the lack of GTL adjustments in the BIOS of any 680i board, regardless of manufacturer. That's because NVidia doesn't allow it, they have it adjusted in the North Bridge. If Intel wanted to harm NVidia, they would have simply pulled their FSB license. People this days have less and less reason to buy a new motherboard. And since NVidia's market for SLI boards is the enthisiast market, which compared to the mass consumer market is pretty small, of course if makes sense for them to throw out there a new chipset and a new motheboard. The reason why we don't see any 780i/790i boards out there yet is because Nvidia is wayting on all the current boards to be sold. So no, I don't think it's bad blood between them and Intel. If anything, I believe that they did this together. The reason why a Dual Core Penryn based chip (Wolfsdale) works in a 680i board, and a Quad Core (Yorksfield) doesn't is because of the GTL adjustments. The fact that there was a Yorksfield ES that worked, whell, let me put it this way: Nvidia didn't want people to stop buying their 680i based boards. That's how business really works. :p Intel doesn't care to go after NVidia for SLI. Of course, I know that it's not what most of you have read or heard, but that's the reality. Intel is a way more flexible company than they used to be. The more plausible theory would be that they will re-enter the discrete graphics market. But keep in mind, that despite popualr belief, this market is not big enough so that everyone can make a good profit margin, that is why so many companies have pulled out of it. Most computer systems that are being sold today worldwide run on a low-end to mid-range Intel CPU on a Intel based motherboard with Intel integrated graphics. Intel is a company that's selling a complete platform solution. Of course, everyone is free to believe as they choose...;)
     
  13. silent-circuit

    silent-circuit [H]ardForum Junkie

    Messages:
    16,133
    Joined:
    Sep 18, 2005
    I'm so glad you informed us -- we, the unwashed masses. We, the proverbially deaf and blind. You, Mr. Wolf, are a godsend.
     
  14. Jodiuh

    Jodiuh [H]ardness Supreme

    Messages:
    5,275
    Joined:
    Feb 3, 2004
    Hopefully this won't. I'm only using a 20in monitor and I've already finished Crysis. I see no need to put up w/ the subpar performance of this chipset and frankly I'm sick of it. Using Winrar makes me want to cry, browsing the net while xferring many files makes me sad, and playing around in the bios bores me. Toss in the lack of Yorkfield support and well, it's time to go back to stability. I'll go play Oblivion and check out Crysis in 3 months. :D
     
  15. RaidalG

    RaidalG [H]Lite

    Messages:
    111
    Joined:
    Jan 15, 2003
    Thanks for the laugh!! I needed that this morning ;)
     
  16. JSpecGC8

    JSpecGC8 Limp Gawd

    Messages:
    416
    Joined:
    Oct 4, 2007


    I am sorry but this is the post of the YEAR!!!

    I am in total agreement, I dont even want to think what would happen if one manufacturer made the best components for an entire computer. Builders across the globe would be hurting as the Budget build would no longer be a budget!!!

    Sorry to say it but GO NVIDIA CRUSH INTEL IN THE GPU MARKET!!!!!
     
  17. Dan_D

    Dan_D [H]ard as it Gets

    Messages:
    53,424
    Joined:
    Feb 9, 2002
    I don't think it was insane at all. Several boards were returned during the early days before NVIDIA fixed the issue via a BIOS update. I was simply impatient and returned those boards about two weeks too early. I used the Striker for a long time but got sick of its' bullshit. The EVGA board may not have the same build quality, but they have been far more reliable while they work. They just haven't lasted very long. Several of the other boards were killed either because I pushed them too hard or due to accidental damage in a couple of cases. Really I've only burned up three of them. Don't forget I've got some friends who have them and they still work just fine. I've also had experience with other 680i SLI boards and I didn't buy them for a number of reasons.

    I simply considered the EVGA the lesser of the evils and kept using those. Plus I wasn't willing to give up SLI support no matter the cost.
     
  18. Dan_D

    Dan_D [H]ard as it Gets

    Messages:
    53,424
    Joined:
    Feb 9, 2002
    I have a feeling that NVIDIA will be fine competing against Intel in that market. In any case I think Intel is a few generations away from being able to challenge them seriously. NVIDIA has always been very aggressive when they've had the need to be.

    Oh that board is fine. The Maximus boards are awesome. I'd be running a Maximus Formula right now if I wasn't so dead set on keeping SLI support.
     
  19. Jodiuh

    Jodiuh [H]ardness Supreme

    Messages:
    5,275
    Joined:
    Feb 3, 2004
    Lose the Dell, wait for Crysis, and you'll be fine! :p

    Fry's put a Maximus/GTS on hold for me...can't wait!! The odd thing's the nipple Max costs less. :confused:
     
  20. STEM

    STEM Gawd

    Messages:
    561
    Joined:
    Jun 7, 2007
    I didn't mean it in a disrespectful or offensive way. It's just that no one else has looked at this issue from this angle. Of course, your sarcasm is greatly appreciated.
     
  21. STEM

    STEM Gawd

    Messages:
    561
    Joined:
    Jun 7, 2007
    The part about "insane" was ment as a joke of course, I never intended to offend you in any way. I hope that the next generation refference model boards will be better. I've read somewhere (it was an interview with someone from Gigabyet) that it would cost an extra $7 per board to use all solid caps. I would gladly pay more for a board to get better quality and reliability.

    Now the advice I was needing from you: Since you've had SLI for a long time, and you get to play with all kind of components, I wanted to ask you what I should do for my next system:

    #1: Go with a Maximus Formula motherboard and 2 X ATI HD3870 in CrossFire - and of course a Q6600 rev. G0

    or

    #2: Go with a ASUS P5N32-E SLI and 2 X EVGA GeForce 8800 GTS 512MB (G92) in SLI - and a Q6600 rev. G0


    I'm just curious if SLI scales better than CrossFire, if driver support is better, and if most games work better. Thanks for your advice in advance.
    :)