DrMrLordX
Lifer
- Apr 27, 2000
- 21,582
- 10,785
- 136
There are somewhat higher failure rates but it isn't a broad systematic issues requiring a recall.
Do we really know that, though? The entire situation could be overblown (hence nVidia's "aww shucks a few units got out of testing, oops" response) or it could be a massive underlying problem with all Turing products that people simply don't yet understand. If 3rd party reviewers could get to the bottom of the issue and explain why the cards are failing, then at least consumers could make an informed choice about whether or not to buy the cards. And of course NV could do the work themselves but as with the 970, they won't.
If we look back at the GTX 970 memory situation it may play out like that. Hearsay, acknowledgement, eventually negligible impact for most people.
The 970 situation was different because reviewers could (and did) figure out the problem on their own. They figured out that memory performance dropped like a rock beyond a certain amount of memory usage, effectively reducing the usable memory of every 970 card. Whether or not the average consumer would know or care about the problem, NV did actually advertise the cards incorrectly, since there was no way to realistically have full use of VRAM. Consumers could examine the data for themselves and make an informed choice accordingly. Sadly it took 3rd party reviewers to uncover the problem.
With RTX cards, we just don't know what's going on, nor do we have usable failure rate data that realistically only retailers and/or NV could provide to us.