Dead 8800 GTX scienced to undeath by oven, mad owner

135

Comments

  • RyderRyder Kalamazoo, Mi Icrontian
    edited December 2009
    If you buy regular Arctic silver, no. If you buy Arctic silver adhesive, yes.

    Regular AS5 will do just fine :)
  • edited December 2009
    Cheers!
  • photodudephotodude Salt Lake, Utah Member
    edited December 2009
    UPSLynx wrote:
    I see what you guys are doing here.

    I'll run some numbers when I get the chance. Probably not until after the weekend.

    still interested in the unbaked vs baked benchmarks.....:bigggrin:
  • UPSLynxUPSLynx :KAPPA: Redwood City, CA Icrontian
    edited December 2009
    yikes, thanks for the reminder!

    I'll put it on the list.
  • edited December 2009
    Could it possibly help a 9600GT that seems to not be working very well and randomly loses signal to the monitor. The card may also be one of the ones with bad bumps on it.
  • UPSLynxUPSLynx :KAPPA: Redwood City, CA Icrontian
    edited December 2009
    Hard to say, but if the problem is intermittent, I'm not sure if baking is the solution for you. You're at a higher risk of losing the card if it still works... my card was completely dead, so I had nothing to lose.

    But who knows, there's only one way to find out.
  • UPSLynxUPSLynx :KAPPA: Redwood City, CA Icrontian
    edited January 2010
    OK, so I benchmarked my post-baked 8800GTX. I found very minor changes in test numbers. Spec Viewperf in 3DS Max showed a loss of 2 points, which isn't a really big deal. Viewperf Maya on the other hand actually showed a slight increase in performance. Cinebench R10 had the same results...

    So uhh.... yep.

    3DSMax Post Bake.JPG
    Maya Post Bake.JPG
    cinebench Post Bake.JPG
  • ThraxThrax 🐌 Austin, TX Icrontian
    edited January 2010
    Summary: No change. These numbers fall within a range that is less than what could be considered the margin of error.
  • photodudephotodude Salt Lake, Utah Member
    edited January 2010
    photodude wrote:
    Did it restore function at the same specs, or could baking cause a performance hit even though it restored function?
    I would expect it functions normally, but you never know...
    Thrax wrote:
    Summary: No change. These numbers fall within a range that is less than what could be considered the margin of error.
    UPSLynx wrote:
    OK, so I benchmarked my post-baked 8800GTX. I found very minor changes in test numbers. Spec Viewperf in 3DS Max showed a loss of 2 points, which isn't a really big deal. Viewperf Maya on the other hand actually showed a slight increase in performance. Cinebench R10 had the same results...

    Good to see that the expected results are as expected, baking restored function at the same specs....within the margin of error
  • ardichokeardichoke Icrontian
    edited January 2010
    I'm thinking that my 8600 may not have come out of the bake quite as well as everyone else's 8800s did. I'm not getting graphic glitches anymore, but gpu folding also doesn't appear to be working. Seems that it always ends up EUE erroring on me now. Haven't tried playing any games since baking my card yet... been too busy with the move.
  • cw
    edited February 2010
    It works,ive tried it myself..
  • EMTEMT Seattle, WA Icrontian
    edited March 2010
    Thanks Lynx, great article. I don't think we've crossed paths but you saved my 8800GTX in addition to your own :)
  • UPSLynxUPSLynx :KAPPA: Redwood City, CA Icrontian
    edited March 2010
    Glad to hear it! I saved my cousin's 8800GT last weekend as well. I'm surprised at how broad the problem seems to be (an how successful the solution has been!)
  • Sledgehammer70Sledgehammer70 California Icrontian
    edited April 2010
    Hey UPS I had a few friends that I directed to this thread & lets just say they are enjoying their baked cards :) I tried it with a 6600 with no luck :(
  • UPSLynxUPSLynx :KAPPA: Redwood City, CA Icrontian
    edited April 2010
    that's great they recovered their cards! (sucks yours didn't make it though :/ ) I'm really surprised at how effective this has been. I haven't heard of many failed bakes since I did mine. Yours and a card that Prime has are the only two I can think of. Who knew?
  • edited April 2010
    I have 2 MSI 8800GTX's OC with the weird lines and stuff. Have been using them as paperweights for about 6 months now. Gonna try baking them this week, will post results :)

    Is yours still working ?
  • UPSLynxUPSLynx :KAPPA: Redwood City, CA Icrontian
    edited April 2010
    My 8800GTX is still working, and I use it quite regularly. No issues at all so far!
  • edited April 2010
    I baked the 2 8800GTX's this weekend, got to try them today, they're both up and running :)
    I have another one so i'll try and get a 780i mobo so I can use them all hehe
  • edited May 2010
    I have seen this method in a couple of places. Your post says to put the gpu facing up some others say the opposite. Facing the gpu down actually makes sense as the solder is on the bottom. Can anyone verify which way they baked thier cards. I have 3 dead 8800 gtx's I would love to bring back to life.
  • ThraxThrax 🐌 Austin, TX Icrontian
    edited May 2010
    Facing the GPU down can also cause it to <i>fall off</i> when the solider is weak. The GPU is not light on this video card.
  • UPSLynxUPSLynx :KAPPA: Redwood City, CA Icrontian
    edited May 2010
    Thrax is right, that is the exact reason why you should face the GPU up.
  • edited May 2010
    Thanks for your responses guys. That makes more sense. Will give it a shot and post back.
  • WingaWinga Mr South Africa Icrontian
    edited May 2010
    Maybe my math is out. Maybe time moves slower in the Southern Hemisphere?
    385F = 196C.... Check
    10minutes northern = 10 minutes southern.... Check

    I just baked a dead 8800gtx and 8 or so minutes into the bake, had electronic components going off like rockets inside the oven.
    One valueble thing I did learn from it though. If you want to harvest a bunch of components off a PC board, overbaking it is the way to go. They pop off with the pins all clean and shiny.

    Im off to scratch in my spares bin to see if I can find another GPU to experiment on.....
  • ThraxThrax 🐌 Austin, TX Icrontian
    edited May 2010
    Are you sure your oven is 100% accurate when you set the temperature?

    Dozens of people have performed this successfully with 385F @ 10 minutes, so it sounds like temperature is to blame.
  • MJOMJO Denmark New
    edited May 2010
    Yes it does sound strange?
    I have baked my card twice, without any problems.
    Even the plastic connectors are unharmed by the heat. :)
  • ardichokeardichoke Icrontian
    edited May 2010
    I actually went out and bought an oven thermometer when I did mine. Good thing too, I Found out my oven at the time ran about 20 degrees hot.
  • WingaWinga Mr South Africa Icrontian
    edited May 2010
    I'm sure I messed something up. I found another card I can try it on. An old AGP X800. Will see how that one makes out. Will have to wait until the week-end though.
  • edited May 2010
    Im going to try this today
  • EMTEMT Seattle, WA Icrontian
    edited July 2010
    Regression! My 8800GTX is showing the same issues it did when I decided to bake it. Should I bake it again guys? Anyone else had to re-bake?
  • ardichokeardichoke Icrontian
    edited July 2010
    Shouldn't hurt to re-bake it... but each time you reflow the solder it's likely not going to bond as well so the fix will last shorter and shorter periods of time. Or at least my limited knowledge of electrical engineering would suggest that to me. You might want to start saving for a new card.
Sign In or Register to comment.