Nvidia settles GeForce GTX 970 memory scandal in the USA

Published by

Click here to post a comment for Nvidia settles GeForce GTX 970 memory scandal in the USA on our message forum
https://forums.guru3d.com/data/avatars/m/227/227994.jpg
Still people write that only 3,5 GB is useable which is not true... but i get the whole point.
https://forums.guru3d.com/data/avatars/m/198/198862.jpg
So, how much are they gonna pay for disabling async compute on Maxwell cards via driver but adverting as "next gen api ready"? Isnt that the same false advertising?
data/avatar/default/avatar19.webp
I guess "overclocker's dream" may be next! Edit: While not quite the same thing it's easy to see how advertising false claims may come back to bite you.
https://forums.guru3d.com/data/avatars/m/54/54823.jpg
Still people write that only 3,5 GB is useable which is not true... but i get the whole point.
It's a technicality, going over has resulted in hitching for some games
https://forums.guru3d.com/data/avatars/m/56/56004.jpg
Let's put it this way, the advertising of the GTX970 as a '4GB' card since 4GB was seen as being a safe amount of VRAM to have ensured its success. Had it been advertised truthfully, perhaps it might not have sold as well. nVidia took a calculated risk, and imo, it paid off. Sure they'd had to pay for the legal fees + $30 for each of those involved in the suit (does that mean ALL GTX970 owners, or just those directly involved in the suit?). Regardless, nVidia'd laughed all the way to the bank, settling this Class action suit is mere pittance to them.
https://forums.guru3d.com/data/avatars/m/258/258664.jpg
Something tells me I never gonna get those $30...
Depends on if you jumped on the bandwagon all these months ago 😉
So, how much are they gonna pay for disabling async compute on Maxwell cards via driver but adverting as "next gen api ready"? Isnt that the same false advertising?
Dude... no it isn't. 'Ready' never means the same as 'fully supporting', for instance, and 'next gen api' doesn't even hold any clear point that one could sue about.
https://forums.guru3d.com/data/avatars/m/235/235224.jpg
You can use both 3.5Gb + the last 512 MB for gaming ? Damn you are reay good as it is seems even Nvidia is not abe to do it.
This topic was beaten to death. You can use the full 4GB, the 512MB was at a slower rate. Looks like the lawyer made a quick buck and even though the payout isn't much for Nvidia, I doubt they'd make the same mistake again after the fiasco. Settling is always better than going through court so it's nice to see they came to an agreement.
https://forums.guru3d.com/data/avatars/m/235/235224.jpg
So slow that unable to be used, and was disabled from the start by Nvidia ( better to have 3.5GB of normal GDDR5 than 3,2 + 712mb of slow ram, because if you wanted to enable the last cache and memory controller of 32b, you had obtain, not even a 3,5GB GDDR pool, but only 3.28GB of full speed GDDR ) Yes it have beaten to death... again and again.... and again and again we see the same bs about it. I personally dont care about it.. http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-970/specifications
Except it is able to be used. G3D, PcPer and others have shown you can use more than 3.5GB, they've shown screenshots of using 4GB. Slow doesn't mean disabled.
data/avatar/default/avatar06.webp
It wouldn't surprise me if the court had referenced G3D forum discussions when deciding the case 🙂 What Nvidia did was bad, both pre and post 3.5GB revelation. But it wasn't something earthshattering, the performance was there, and it has been accurately measured. And the settlement reflects this.
https://forums.guru3d.com/data/avatars/m/240/240526.jpg
These class actions are pointless. The only real winners here are the lawyers on the winning side. They will no doubt get millions.
https://forums.guru3d.com/data/avatars/m/80/80129.jpg
These class actions are pointless. The only real winners here are the lawyers on the winning side. They will no doubt get millions.
The point is that it's a penalty to the company so they don't do again in the future. Not for people who bought the card to get paid.
https://forums.guru3d.com/data/avatars/m/191/191875.jpg
thats peanuts compared to the revenue they made out of selling the cards as 4GB cards.
Oh please no one was buying these cards because they had 4Gb of Vram they bought them because the benchmarks for the cards showed it was a fantastic price / performance powerhouse, a factor that stayed the same throughout the whole 4Gb non sense, the only reason this got any traction is too many people with to much free time loving to complain about **** that doesn't matter.
https://forums.guru3d.com/data/avatars/m/80/80129.jpg
Oh please no one was buying these cards because they had 4Gb of Vram they bought them because the benchmarks for the cards showed it was a fantastic price / performance powerhouse, a factor that stayed the same throughout the whole 4Gb non sense, the only reason this got any traction is too many people with to much free time loving to complain about **** that doesn't matter.
I wouldn't say it doesn't matter. Nvidia should definitely be penalized for it, regardless to whether it was intentional or not. That being said, I agree with you that it was mostly overblown. The vast majority of people look at the initial benchmarks and buy the cards, they don't scrutinize memory configurations or architecture faults. Had Nvidia explained the memory situation for the cards launch, I doubt anyone would have cared.
https://forums.guru3d.com/data/avatars/m/123/123440.jpg
You need to make a correction in the article "These cards also had 56 instead of 64 rops, while the company initially advertised 56 initially" should say 64 initially?
https://forums.guru3d.com/data/avatars/m/180/180832.jpg
Moderator
Something tells me I never gonna get those $30...
yes you would need to emigrate to USA, rest of the world no chance
https://forums.guru3d.com/data/avatars/m/115/115462.jpg
Oh please no one was buying these cards because they had 4Gb of Vram they bought them because the benchmarks for the cards showed it was a fantastic price / performance powerhouse, a factor that stayed the same throughout the whole 4Gb non sense, the only reason this got any traction is too many people with to much free time loving to complain about **** that doesn't matter.
I can add that I had 2x970s for almost 2 years. Sure I had the SLI related issues over time, but I passed 3.5GB mark loads of times, and I never noticed anything, because of it (I'm not saying there wouldn't have been any problems, I simply didn't notice them). The only times where vram stuttering was present was when I had over 3.9GB usage or so, where the card was obviously running out of physical vram. Anyway, I wouldn't mind getting 60$, but I won't for sure so yeah, I can only hope nvidia learned a lesson or two from this. If anything, the biggest damage is their image, less so the money which is negligible compared to their profit.
https://forums.guru3d.com/data/avatars/m/144/144047.jpg
i doubt ill see that money lol,still lovin my 970 anyway
https://forums.guru3d.com/data/avatars/m/196/196284.jpg
I wouldn't say it doesn't matter. Nvidia should definitely be penalized for it, regardless to whether it was intentional or not. That being said, I agree with you that it was mostly overblown. The vast majority of people look at the initial benchmarks and buy the cards, they don't scrutinize memory configurations or architecture faults. Had Nvidia explained the memory situation for the cards launch, I doubt anyone would have cared.
I got the card based solely on it's performance. Memory configuration was never a consideration and had it been advertised accurately, I'd have still bought the card. If I was a gamer, I'd even buy a second for SLI without hesitation.
https://forums.guru3d.com/data/avatars/m/235/235224.jpg
Gimped settlement for a garbage gimped product.. I wish the users got more than 30 bucks.
Garbage product?