Discussion in 'Videocards - NVIDIA GeForce Drivers Section' started by newls1, Mar 27, 2018.
i be very happy if i see 1 fps improvement by every driver released
Although 1fps isn't much, if every driver released had a 1fps improvement, then 5 drivers later we'd be up by 5fps. That's mostly wishful thinking though. Seriously though anyone try final fantasy with these?
Today was patch for bf
Vega56 still has a 10% advantage ahead of the 1070.
It's called margin of error...
yeah... i have these margin of error in every last 5 or 6 drivers... finally i have margin of error to more and not less
That's...not how it works.
A. Vega 56 is about 10% ahead of 1070 on average in all games.
B. FC5 is AMD's game so whatever was optimized in it wasn't optimized for NV h/w.
you can read?
391.35 - r391_33-2
definitely r391 is different branch than r390
I haven't ran smooth since the affliction crap came out.
It's pretty obvious who can't read here.
looks you really have one problem to accept when you say anything clearly wrong
that´s so stupid
but the fact are just one... are r391 branch and not r390
just a quick confirmation that the firefox TDRs are still not fixed in 391.35
Yeah, I'm clearly wrong, as well as the company which made the driver. You're the only one right and it's me who's having a problem here.
company? the company released a driver using branch r391_33-2
just need read this to know is different than r390 or to be more specific the r390_99 or r390_00 range
or you have a problem or don´t have knowledge about what you´re discussing here
Branch is clearly stated in the release notes. Internal version of the driver which you refer to is not always indicative of the branch its from.
I honestly have nothing more to add. This driver is from R390 branch, same branch as all previously released 390.xx drivers. If you want to prove otherwise I'd suggest you go directly to Nvidia since it's them who is stating this, not me. Good luck.
hahahaha ok if driver says r391 but one pdf call him of r200 so are r200 branch driver
and same branch than olders 390xx drivers? you really don´t know what you saying
Driver 391.35 is r391_33-2
Driver 391.24 is r390_99-9
Driver 391.05 is r390_99-5
Driver 391.01 is r390_99-2
Driver 390.77 is r390_00-154
yeah... looks the same hahahaha
Yeah, the Spectre protection does make BF1 stutter once in a while - I think it's because BF1 is such a CPU heavy game, and the Spectre protection makes CPUs less efficient. I average about about 65% Total CPU Usage on my 4.6Ghz 6700K, with spikes to 100% on all cores, according to HWInfo, (that's on Team Deathmatch with only 20 players though, I haven't done measurements on the 64 player Operations Mode - CPU usage probably even higher there). It's possible that BF1 might be smoother with CPU's with 6 or more real cores with this Spectre protection enabled, because there's not much/any overhead left when running BF1 on a 4c/8t CPU. (I am talking 120-144fps though, so that increases CPU usage too).
A. Vega 56 is about 3% faster in FHD and 7% faster in 1440P, so overall it's a bit bigger gap than the average. But if you check other benchmarks in 1440P: TPUp 49-61 fps, Techspot 64-79, Kitguru 66-82. All 3 is about 25% difference, not 10-11%. PCGamer also has a 10% difference. Anyway, the 10-11% difference is bigger than the average difference.
B. I leave this here to taste the difference what an AMD and NV sponsored title looks like:
Far Cry 5 as an AMD sponsored game: running decent on both manufacturers' cards.
Far Cry 4 as an NV sponsored game: "Ubisoft continued its partnership with nVidia into Far Cry 4, featuring inclusion of soft shadows, HBAO+, fine-tuned god rays and lighting FX, and other GameWorks-enabled technologies. Perhaps in tow of this partnership, we found AMD cards suffered substantially with Far Cry 4 on PC. "
Far Cry 4 also ran better on AMD h/w during its release (I think that it was only one GW feature which were too heavy for GCN cards but eh, whatever). Far Cry 5 runs worse on NV h/w on every setting option. So yeah, "taste the difference".