It’s not about defending Blizzard… if plenty brands & models were also affected by sudden massive over usual statistics values GPU RIP, then we may indeed begin to question ourselves about Diablo IV’s responsability. But if the issue is statistically over-affecting a specific brand/model, then the chance it’s a bad component issue used for a series of cards or poor hardware design issue (not counting the possibility of user’s responsability in case of unsupported overclocking, having tons of dust, etc.).
Hardware manufacturers are supposed to design solid & reliable security systems (for sure, it never excludes accidents), and every piece of high priced hardware is supposed to pass quality assurance tests before leaving to manufactury. Demanding softwares should never cause harm to GPU (it’s not like a malware designed to continuously write data on an HDD/SDD which has a limited number of write cycles limits & which may finally really cause hardware damage).
For sure a GPU has mecanical pieces too (fans), but they are supposed to be designed to work at 100% for very long, and again there should be a robust hardware security system causing the GPU to enter into throttle mode in case suddenly the temperature begins to get too hot comparated to the max. temp. it’s supposed to handle.
However, now there are enough reports about such a specific issue, Blizzard has to react, to avoid having more casualties, for example in the first stage, detecting the model of the installed GPU & if possible the manufacturer (GPU-Z can do that but seems to require administrator right to work, but I don’t know if admin is needed for that specific feature or for another reason), and if the potential deadly combo is detect, take some action : either refuse to launchs the game or force a very reduced performance (e.g. very low fps/reduced quality/etc.) for these cards at the “potentially dangerous places” and probably work with Nvidia and/or Gygabyte to find a better solution on the long term.