3 ms·

Isn't this reasonable though? Surely if I overclock and burn out my GPU, NVidia shouldn't be on the hook?

wtallis 9y ago
The hardware shouldn't be relying on drivers in the host OS to protect it from melting. That's too failure-prone even when the user isn't trying to push the limits of the hardware through overclocking. Thermal and voltage limits need to be enforced by the peripheral device itself either in hardware or in its own firmware. Relying on software running on the host CPU to provide a realtime response to critical situations is foolish. Extreme overclocking of the sort that can break your hardware should require soldering or external re-flashing of a firmware ROM.
- mulmen 9y ago
  What if I change the firmware? What if I change the hardware? Where is the line and why distinguish between a pure hardware product and a combined product that uses both hardware and software? Why should we legislate engineering decisions?
  - wtallis 9y ago
    I don't think any of those are productive questions. If you alter firmware that is usually unchanged over the life of the product (as for a desktop graphics card) and you specifically edit the rules it contains for thermal throttling, then you are obviously liable for the lump of slag that you end up with. Likewise if you're making hardware modifications that compromise the thermal management capabilities. In either case, it will be pretty easy to show after the fact that the product was modified, and that your modifications were probably related to the failure. None of that applies when the host software is responsible for preventing hardware damage. As I've explained, relying on software running inside of Windows to prevent hardware damage is asking for trouble even without user intervention. You may also notice that I'm not advocating for "legislating engineering decisions"; though there are plenty of situations where engineering decisions obviously should be determined by legislation, this is more a question of after the fact liability to be worked out by the courts, which are generally pretty good at figuring out whether to place the blame on a user or on the engineers.
  - bradstewart 9y ago
    It all boils down to personal responsibility which, admittedly, doesn't seem to be a particularly valued trait these days. Both you and the manufacturer know if your customization was the likely cause of a failure or not. Life would be much simpler if everyone just admitted that and moved on.
- function_seven 9y ago
  Reminds me of Dell denying warranty support due to VLC damaging the speaker https://hardware.slashdot.org/story/14/02/09/1828229/customer-dell-denies-speaker-repair-under-warranty-blames-vlc https://hardware.slashdot.org/story/14/02/09/1828229/custome...
YetAnotherNick 9y ago
I was referring to jailbreaking my iPhone.
- manfredo 9y ago
  Say you jailbreak your phone and the new operating system has bad state of charge tracking software. Should Apple be on the hook for a new battery? I think that a lot of people who agree with your statement aren't considering that running third party software on a device might inflict costly (or irreparable) damage on a product.
  - wtallis 9y ago
    Isn't battery charge management handled by an entirely separate microcontroller that functions even when the application processor is powered down? I know that's how Apple's laptops have always worked.
  - stale2002 9y ago
    IF the bad software causes the failure, then no Apple should not be in the hook. But if it is not, then yes they they should be. And the burden of proof that says that the software caused the issue, should be in Apple.

The hardware shouldn't be relying on drivers in the host OS to protect it from melting. That's too failure-prone even when the user isn't trying to push the limits of the hardware through overclocking. Thermal and voltage limits need to be enforced by the peripheral device itself either in hardware or in its own firmware. Relying on software running on the host CPU to provide a realtime response to critical situations is foolish. Extreme overclocking of the sort that can break your hardware should require soldering or external re-flashing of a firmware ROM.

mulmen 9y ago
What if I change the firmware? What if I change the hardware? Where is the line and why distinguish between a pure hardware product and a combined product that uses both hardware and software? Why should we legislate engineering decisions?
- wtallis 9y ago
  I don't think any of those are productive questions. If you alter firmware that is usually unchanged over the life of the product (as for a desktop graphics card) and you specifically edit the rules it contains for thermal throttling, then you are obviously liable for the lump of slag that you end up with. Likewise if you're making hardware modifications that compromise the thermal management capabilities. In either case, it will be pretty easy to show after the fact that the product was modified, and that your modifications were probably related to the failure. None of that applies when the host software is responsible for preventing hardware damage. As I've explained, relying on software running inside of Windows to prevent hardware damage is asking for trouble even without user intervention. You may also notice that I'm not advocating for "legislating engineering decisions"; though there are plenty of situations where engineering decisions obviously should be determined by legislation, this is more a question of after the fact liability to be worked out by the courts, which are generally pretty good at figuring out whether to place the blame on a user or on the engineers.
- bradstewart 9y ago
  It all boils down to personal responsibility which, admittedly, doesn't seem to be a particularly valued trait these days. Both you and the manufacturer know if your customization was the likely cause of a failure or not. Life would be much simpler if everyone just admitted that and moved on.
function_seven 9y ago
Reminds me of Dell denying warranty support due to VLC damaging the speaker https://hardware.slashdot.org/story/14/02/09/1828229/customer-dell-denies-speaker-repair-under-warranty-blames-vlc https://hardware.slashdot.org/story/14/02/09/1828229/custome...

YetAnotherNick 9y ago

I was referring to jailbreaking my iPhone.

manfredo 9y ago
Say you jailbreak your phone and the new operating system has bad state of charge tracking software. Should Apple be on the hook for a new battery? I think that a lot of people who agree with your statement aren't considering that running third party software on a device might inflict costly (or irreparable) damage on a product.
- wtallis 9y ago
  Isn't battery charge management handled by an entirely separate microcontroller that functions even when the application processor is powered down? I know that's how Apple's laptops have always worked.
- stale2002 9y ago
  IF the bad software causes the failure, then no Apple should not be in the hook. But if it is not, then yes they they should be. And the burden of proof that says that the software caused the issue, should be in Apple.