Report: All Nvidia G84/G86 Chips Defective

[JerryC]

Your lucky, I been through two 8800gt's in the last 6 months.


[citation][nom]TheDozer42[/nom]I've been hammering away at my 8800 for a year and a half, no problems whatsoever.[/citation]

 

[Guest]

Oh my God.. what happens if over a few million users in the world got affected? I would love to see Nvidia lose to ATi and we get our refund.. my 8600GT still works fine for now

 

[modtech]

this is what's sad... you are using aftermarket cooling and ur 8800gt runs at 52c. My stock 8800GTS OC 512 has no aftermarket cooling and runs at 52c.

G92's and 94's should not be a affected.

By G84 and G86 they mean that chipset style (found in 8400-8600 chips both desktop and laptop).

I have only had Nvidia cards, my last 6600GT lasted me 4 years and I'm already on 7 months with my new 8800GTS OC and its been great.

Just a streak of bad luck for Nvidia... ATI has it ALL THE TIME!

Dude, that is 52c on FULL LOAD with fan speed at 1/2 (It's inaudible) AND it's overclocked. Your 8800GTS OC is nowhere near as silent as mine. In case you're wondering, yes the rest of my rig is also silent, I can sleep right next to it.

Stock cooling on the 8800GT was so bad, that it would sometimes hit 70c idle unless one manually forced the fan to spin higher. Dumbass.

 

[forumhound]

Well I will never buy another nvidea chip again, after two dead logic boards in a mac book pro and the apple dealer just shaking their heads. hundreds of others have already posted failures of MBPs and all apple can do is replace the board with another one that will go bad in x amount of time. a total fiasco. glad the Enquirer is on this, and John Edwards. anyway, i lost so much time/work/money on that chip, i am totally sour on nvidea now. cheers!
FH

 

[howardp6]

My PNY 8600 GT keeps over heating, even after replacing my Case with a Thermaltake case, adding four case fans, setting the fan to full speed with Oblivion and Mass Effect. The only after setting the fan to full speed on 3D applications and downclocking the GPU and video memmory could I get it not to show artifacts and the system locking up. Could this be caused by the defective G86 chipsets. I replaced the PNY 8600 GT with an ATI 4850 in the same case and have had no issues.

 

[forumhound]

@Howardp6, perhaps not. At least in macbookpros, the chip will start to fail intermittently for a bit then go completely dead forever. it may be caused by heat who knows, but the chip is either on or off in each case.

 

[Guest]

Please at least read Charlie's recent series of articles - throughly explained and researched. The problem is not due to overheating. It will still occur under normal temperatures, it's just that it will take longer to manifest itself. The problem is nothing to do with inadequate cooling, but the manufactuing process of the chip itself.

I don't blame people for reacting (flaming) the way they are - denial is much more preferable to worrying that your 8800 could blow at any point.

Would you really expect nVidia to come out and say their video cards are doomed to failure? That would be business suicide - shareholders jump ship. The failures will happen over a long period of time so no one will notice - and as Charlie points out, the failure might not happen within the life of the chip anyway - it depends mostly on the thermal cycles the chip undergoes, that is why laptops are going first.

Time will tell - meanwhile, I'll stick with ATI, thanks.

 

[apache_lives]

update: my workplace got 3 dud 9800GTX+ cards in a week (2 gigabyte, 1 asus) - they worked for an hour to install windows etc then packed it - tell me what this means...

 

[benvanderjagt]

If anyone's still looking up this thread, I'd like to throw in my 2c worth. I've been reballing and reflowing ATi, nVidia, and Intel chips for quite some time now. The nVidia g86 chips are failing at about the same rate as the 7600 chips. While I believe the liquefying underfill -could- cause chips to die, the vast (greater than 90%) majority of failures is STRICTLY BGA! I remove, reball, and reattach the chips, and they work. I think it's the lead-free balls that cause the problems. If the underfill were truly the problem, then wouldn't reflowing the chips have zero effect? If the problem were a defect in nVidia g86 chips, then why are the 7600 chips dying just as fast?

 

[apache_lives]

woa old thread haha

7 series i can confirm (confirm aka dead -> reflow -> work) has the same issues is the 8 series and even the 9 series and all chipsets from around these generations

you can see the after effect from this event in 07-08 -- hp's mainstream laptops (DV6 etc) use ATI video now rather then nvidia

 

[benvanderjagt]

I am now reballing ATi chips at almost the same frequency as nVidia chips, so I sincerely think the failures have to do with the lead-free solder balls. I talked with a pro who is much more experienced than I am, and he had the same suspicion, saying that the lead-free balls just aren't up to the task. I'm indeed very happy to see AMD and Intel both putting their GPU's on the CPU, something of a holy grail in the industry for a decade or so.

I also am seeing plenty of Intel northbridge and southbridge failures, but generally I don't fix the northbridges, because they usually have many, many balls and cause me to swear a lot. (-: