Almost certainly means electronics or semiconductors burnt up on the board. Not salvageable without PC board repair.
:-(((
Indeed, a stuck fan wouldn't prevent bootup. It's actually stopping the motherboard from powering on, but 5V standby is present on the motherboard. I'm thinking the 12V input to the graphics card (those extra power connectors on the top) are shorted in the voltage regulator, and the main PSU is detecting this and refusing to turn on. The PSU smells a bit off too, but it seems to be working fine, it's a decent high end Corsair, maybe it tripped just in time to save itself. Broken graphics cards seem to sell for half the price of working ones on Ebay, very odd since someone repairing them would make a good one out of about 2 broken ones, so make no profit at all. Maybe they can make 4 good ones out of 5 if there are enough different failure points, or even ones that can be fixed at little or no cost.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
You would have to be already set up with a hot-air workstation for reworking SMD mounted parts and have experience doing PC board repair including circuit board trace repair and tunneling. Pretty extensive rework skills and a lot of time involved. Might not be cost effective for the amount of time spent doing the rework depending on the cost of your labor of course.
You would have to be already set up with a hot-air workstation for reworking SMD mounted parts and have experience doing PC board repair including circuit board trace repair and tunneling. Pretty extensive rework skills and a lot of time involved. Might not be cost effective for the amount of time spent doing the rework depending on the cost of your labor of course.
I'm wondering if the people buying up the broken ones are just remounting the cooler and reflowing solder by sticking the whole thing in an oven - as most faults are due to an overheat causing some of the GPU's solder contacts to come off? In which case mine may not sell. I could lie and say "it just stopped working", but I'm too honest for that, and don't want it being sent back either!
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
Bad smell means some part of the pcb substrate burned up or severely overheated. The bad smell is from the organic based polymers that are using in making the semiconductors.
If you are lucky only one or two devices burned up without destroying or lifting the solder pads on the PCB. That can be fixed with a new part installed. Worse if the vias or internal traces in the PCB burned up. Worse still if the PCB material is burned up or scorched. That involves dremeling out all the burned substrate and rebuilding the traces and replacing the material.
I have many years of experience in PCB repair. I've spent weeks repairing and troubleshooting a single PCB only because the replacement cost of a new PCB was $20,000. Made my time worthwhile.
Bad smell means some part of the pcb substrate burned up or severely overheated. The bad smell is from the organic based polymers that are using in making the semiconductors.
If you are lucky only one or two devices burned up without destroying or lifting the solder pads on the PCB. That can be fixed with a new part installed. Worse if the vias or internal traces in the PCB burned up. Worse still if the PCB material is burned up or scorched. That involves dremeling out all the burned substrate and rebuilding the traces and replacing the material.
I have many years of experience in PCB repair. I've spent weeks repairing and troubleshooting a single PCB only because the replacement cost of a new PCB was $20,000. Made my time worthwhile.
So a bad smell could be PCB or component damage? Well it was a mildly bad smell, only just detectable by sniffing closely. I've had smells where the whole room stinks.
According to Ebay buyers it's maybe not fixable, since 50 folk have looked but not bought.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
One of the major benefits of "multi-threading" when one or more of the threads is paused (like for memory access or other I/O) the other threads are continuing processing of that application. This assumes "asynchronous processing" however.
If the threads are synchronously processed (and can't get out of step at all) there is usually no visible qualitative advantage, just the linear scale up of more production per thread.
This asynchronous processing allows a multi-threaded app to process anywhere from "somewhat faster" to "incredibly faster".
I remember a report on a Ibm 370 mainframe ("several" years ago) where they changed the JCL on a batch application that had sub-routines (the sub-routines didn't have a dependency on mainline) to be "asynchronous" and the result was a 40% decrease in total run time (the cpu was a lot busier too :) They were very happy because it was no longer taking the entire night shift to run that production job.
You get the same effect when you run multiple apps on a GPU. The "overlapped" processing and/or I/O allows you to get a some speedup compared to a single task on the GPU (assuming sufficient memory and good architecture etc).
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Once you've burned up a few boards and various components, I could tell just by the smell whether it was a composite resistor, transistor, IC, SMD component or board substrate. Like you did, one of my first troubleshooting steps was just to smell the board closely. Then eyeball obvious scorched, melted or overheated components. First replace those parts and then figure out why they failed in the first place.
Once you've burned up a few boards and various components, I could tell just by the smell whether it was a composite resistor, transistor, IC, SMD component or board substrate. Like you did, one of my first troubleshooting steps was just to smell the board closely. Then eyeball obvious scorched, melted or overheated components. First replace those parts and then figure out why they failed in the first place.
I can't tell by the smell, even if I was good at that sort of thing, the smell wasn't particularly strong. Perhaps that's a good thing - only one thing could have blown. It does seem strange though, if it's managing to short circuit an 850W Corsair PSU, wouldn't there be more black and smelly stuff? Or could it be shorting the 12V line on the PCI Express socket? I shall remove the heatsinks tomorrow to see if I can find the dead component(s). Since those cards have shot up in price recently (in fact I think all GPUs have - coronavirus - everyone playing games at home?), it may be worth buying some parts to fix it.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
Most good power supplies will crowbar their outputs if they detect trying to drive a dead short. So that is why the power supply won't show any output with the cards plugged in. Probably why there was little smell, the card didn't get supplied +12V long enough to really burn things up.
Keith Myers wrote: Quote:It
)
:-(((
mikey wrote: Keith Myers
)
Indeed, a stuck fan wouldn't prevent bootup. It's actually stopping the motherboard from powering on, but 5V standby is present on the motherboard. I'm thinking the 12V input to the graphics card (those extra power connectors on the top) are shorted in the voltage regulator, and the main PSU is detecting this and refusing to turn on. The PSU smells a bit off too, but it seems to be working fine, it's a decent high end Corsair, maybe it tripped just in time to save itself. Broken graphics cards seem to sell for half the price of working ones on Ebay, very odd since someone repairing them would make a good one out of about 2 broken ones, so make no profit at all. Maybe they can make 4 good ones out of 5 if there are enough different failure points, or even ones that can be fixed at little or no cost.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
You would have to be already
)
You would have to be already set up with a hot-air workstation for reworking SMD mounted parts and have experience doing PC board repair including circuit board trace repair and tunneling. Pretty extensive rework skills and a lot of time involved. Might not be cost effective for the amount of time spent doing the rework depending on the cost of your labor of course.
Keith Myers wrote: You would
)
I'm wondering if the people buying up the broken ones are just remounting the cooler and reflowing solder by sticking the whole thing in an oven - as most faults are due to an overheat causing some of the GPU's solder contacts to come off? In which case mine may not sell. I could lie and say "it just stopped working", but I'm too honest for that, and don't want it being sent back either!
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
Bad smell means some part or
)
Bad smell means some part of the pcb substrate burned up or severely overheated. The bad smell is from the organic based polymers that are using in making the semiconductors.
If you are lucky only one or two devices burned up without destroying or lifting the solder pads on the PCB. That can be fixed with a new part installed. Worse if the vias or internal traces in the PCB burned up. Worse still if the PCB material is burned up or scorched. That involves dremeling out all the burned substrate and rebuilding the traces and replacing the material.
I have many years of experience in PCB repair. I've spent weeks repairing and troubleshooting a single PCB only because the replacement cost of a new PCB was $20,000. Made my time worthwhile.
Keith Myers wrote:Bad smell
)
So a bad smell could be PCB or component damage? Well it was a mildly bad smell, only just detectable by sniffing closely. I've had smells where the whole room stinks.
According to Ebay buyers it's maybe not fixable, since 50 folk have looked but not bought.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
One of the major benefits of
)
One of the major benefits of "multi-threading" when one or more of the threads is paused (like for memory access or other I/O) the other threads are continuing processing of that application. This assumes "asynchronous processing" however.
If the threads are synchronously processed (and can't get out of step at all) there is usually no visible qualitative advantage, just the linear scale up of more production per thread.
This asynchronous processing allows a multi-threaded app to process anywhere from "somewhat faster" to "incredibly faster".
I remember a report on a Ibm 370 mainframe ("several" years ago) where they changed the JCL on a batch application that had sub-routines (the sub-routines didn't have a dependency on mainline) to be "asynchronous" and the result was a 40% decrease in total run time (the cpu was a lot busier too :) They were very happy because it was no longer taking the entire night shift to run that production job.
You get the same effect when you run multiple apps on a GPU. The "overlapped" processing and/or I/O allows you to get a some speedup compared to a single task on the GPU (assuming sufficient memory and good architecture etc).
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Once you've burned up a few
)
Once you've burned up a few boards and various components, I could tell just by the smell whether it was a composite resistor, transistor, IC, SMD component or board substrate. Like you did, one of my first troubleshooting steps was just to smell the board closely. Then eyeball obvious scorched, melted or overheated components. First replace those parts and then figure out why they failed in the first place.
Keith Myers wrote: Once
)
I can't tell by the smell, even if I was good at that sort of thing, the smell wasn't particularly strong. Perhaps that's a good thing - only one thing could have blown. It does seem strange though, if it's managing to short circuit an 850W Corsair PSU, wouldn't there be more black and smelly stuff? Or could it be shorting the 12V line on the PCI Express socket? I shall remove the heatsinks tomorrow to see if I can find the dead component(s). Since those cards have shot up in price recently (in fact I think all GPUs have - coronavirus - everyone playing games at home?), it may be worth buying some parts to fix it.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
Most good power supplies will
)
Most good power supplies will crowbar their outputs if they detect trying to drive a dead short. So that is why the power supply won't show any output with the cards plugged in. Probably why there was little smell, the card didn't get supplied +12V long enough to really burn things up.