Discussion in 'Videocards - NVIDIA GeForce Drivers Section' started by robalm, Oct 7, 2015.
Not yet. 385.8x or 9x will be if won't be any delay ofc.
worse performance and stability than 356.04, uninstalling...
Was going to ask this ^^^
Anyone else compared these 358.50 WHQL with 356.04 Dev?
Not gonna have enough time to do some cross testing between two of these until weekend, also I'm not willing to dump rock solid 356.04 Dev if 358.50 WHQL are not good at least like 356.04 are.
Have an issue with DSR and my 4KTV with these. On previous drivers, it would use 1080p as the native resolution and increase DSR based on that. With these, even though it says that 1080p is the native resolution, DSR scaling starts scaling up from 4096 x 2160, which is obviously undesirable.
Back to 356.04 for me.
So is the glitching F.E.A.R. shadow finally fixed in this driver? I am guessing not...
Is this driver from a new branch then? If so then I wonder what NVIDIA have managed to break in this set?
Some additional info, potentially interesting for SLI and Optimus users in Win10:
Wow, you've tested the overall performance and stability on all systems on the planet in like two hours from the moment of the driver release.
I'll try these on win10, drivers other than 356.04 throw up a lot of driver recovered errors. Hope these are fine.
Valley Benchmark Windows 10 x64
FF XIV DX11 bechmark shows same score as before but look at Valley above it's horrible. Something is really wrong with these drivers on my system.
A loss of 1000+ is not normal and I never saw anything like that. Same setting same NVCP settings nothing was altered.
Keep flamebaiting and you'll be getting infractions on your account
Are you sure you ran the same valley test? Those results don't make any sense otherwise.
FEAR has been fixed in these drivers! Happy days!
You seem angry. That's not good!
Also, a 560 Ti is roughly about as powerful as a 750 Ti, a low end, but still pretty capable card.
Yes same settings. I did nothing to the bechmark itself nor to the NVCP. I'm still scratching my head to find what's wrong. It makes no sense to me too.
U will be surprised with the performance of fermi gpu cards in direct x 12.
Fermi has Hardware scheduler and kepler software.
It seems nvidia is crushing the 700 series performance. I'm not sure if the 900 series suffer from the performance downgrades but this doesn't look good when pascal comes.
I believe nvidia will cripple maxwell just as they're crippling kepler
I'm pretty sure Intel doesn't hurt they chips and degrade performance since the performance lasts for years before a upgrade.
If amd had proper drivers, it would've sorted nvidia out and fix whatever is happening with these drivers and why it's making performance worse with every new release
^ In an economical world, this is normal...
Just tested the driver.
1.- Pinball FX2 still a stuttering disaster with super sampling.
2.- Mortal kombat X still jittery, might be better but I still need to run stream flash video on second plane for it to play properly smooth.
3.- F1 2015. Performance tanks when restarting a quick race, doesn´t happen on 353.49. Didn´t play long enough to test for the ctd video memory error.
4.- Soma, still plays horribly. Good fps but smoothness isn´t there.
6.- Tried to test NBA 2K15 for stuttering, but it crashes to desktop instantly 5 out of 5 times. Installed 353.49 and crashes are gone. Unplayable because of stuttering of course.
5.- Installed whole driver package by mistake and lost audio in all of my games.
6.- Lost 20 minutes for nothing.
It´s still Kepler and we know high end cards aren´t that far away from the lower model with overclocking.
As much as it is, nvidia pretty much takes full advantage of this. I skipped maxwell for pascal when it releases but this is ridiculous with the downgrades.
We're talking about benchmarks and games that performed great only to progressively perform worse with each update.
I'm getting to old anyway and have a son to raise. I guess it may be a good idea to hang in the hat. Don't think I may upgrade if this is nvidia grand roadmap to success..
I've spent enough
A lot of tests were made around this claim. None of them proved it.
Lack of driver optimization for a previous architecture h/w is something that is expected from every IHV. Doesn't make these cards slower, just makes them not as fast as they could be in new titles.
I don't mind waiting some more, but it would have been nice to run a few DX12 benchmarks prior to launch...