https://forums.geforce.com/default/topic/951723/geforce-drivers/announcing-geforce-hotfix-driver-368-95/
>>55708645
doesn't seem to completely get rid of the latency but it's a start
>>55708645
A GeForce driver is an incredibly complex piece of software, We have an army of software engineers constantly adding features and fixing bugs. These changes are checked into the main driver branches, which are eventually run through a massive QA process and released.
>army of software engineers
>ARMY
>>55708645
>doesn't fix anything
so AMD can fix a power draw issue and NVIDIA can't even fix stuttering and DPC latency? and AMD are supposed to have bad drivers?
>>55709961
>power draw fix
Literally downclocking gpu
>>55709987
except that was an optional feature of it, the main, unchangeable fix did not downclock it
>>55709911
I've had this shit happen on my GTX970.
What I first thought to be a Nvidia Driver because that was what LatencyMon reported turned out to be an wifi driver issue, changed to wired and problem solved.
>https://forums.geforce.com/default/topic/951723/geforce-drivers/announcing-geforce-hotfix-driver-368-95/
>https://forums.geforce.com/default/topic/941579/geforce-1000-series/gtx-1080-high-dpc-latency-and-stuttering/
>http://www.overclock.net/t/1605618/nv-pascal-latency-issues-hotfix-driver-now-available
also have you noticed any of the screen flicking that a lot of pascal users have been enduring for two months now?
>https://forums.geforce.com/default/topic/939358/geforce-1000-series/gtx-1080-flickering-issue/
>>55710047
DELETE THIS
>>55710047
>>55710000
Checked
>>55710047
>>https://forums.geforce.com/default/topic/939358/geforce-1000-series/gtx-1080-flickering-issue/
i have flickering all the time on my brand new zotac gtx 1060, also a lot of microstutters
>>55710055
https://www.youtube.com/watch?v=wLR5u_Tw8lA
>>55708645
>no async fix
>no vulkan fix
>no DX12 fix
ok
>>55710000
>>55710047
Good thing I'll wait till next month to buy my new PC
Thank you for beta testing
bump for fermi justice
>>55708770
>>55709961
>>55709987
>>55710154
NVIDIA FAGS BTFO!!! HAHA ENJOY YOUR SHIT CARDS!!
AMD ALL THE WAY!
nVidia is "fixing" DPC-issues for years now.
It's embarrassing.
>>55708645
haven't noticed any latency with my 1080.
>>55713124
Because you're a tech-illiterate retard who has no idea what DPC latency is. Go back to /v/.
I had the problem solved with hotfix.
anyway, I'm nvidiot since always and this is really the first time that such shit happened. don't be dicks. these guys are masters in drivers.
DO YOU KNOW IF nvidia drivers are universal? like if I switch for example to 650, 970, 1070....are all "reasonably modern" gpus supported?
>>55713227
>this is really the first time that such shit happened. don't be dicks. these guys are masters in drivers
http://wccftech.com/nvidias-latest-game-ready-driver-allegedly-killing-gpus-plagued-issues/
http://www.theinquirer.net/inquirer/news/2450200/nvidia-geforce-36447-gpu-driver-is-borking-gamers-pcs
http://www.lazygamer.net/gaming-news/pc-gaming/be-warnednvidias-latest-driver-is-destroying-some-graphics-cards/
http://modcrash.com/nvidia-display-driver-damaging-gpus/
http://www.zdnet.com/article/warning-nvidia-196-75-drivers-can-kill-your-graphics-card/
http://fudzilla.com/news/graphics/40364-nvidia-geforce-364-72-drivers-come-with-plenty-of-issues
http://www.pcauthority.com.au/News/417802,gefail-nvidia8217s-driver-fiasco.aspx
In before the inevitable defensive response of a list of AMD driver issues, as if that excuses everything.
I've had no issues with my GTX 1080, everything runs super smooth and my temps stay below 70 degrees unlike shitty AMD cards.
>>55712574
if only there were any decent freesync monitors.
no 240hz for us. :(
http://www.144hzmonitors.com/monitors/asus-rog-swift-pg258q-24-5-inch-full-hd-g-sync-240hz-gaming-monitor/
sad shit.
This is WITH hotfix on my 1070 GTX.
>>55713176
you got me
>>55713829
AMD cards work on my machine :^)
It's good new that Nvidia fixed something through drivers instead of gimping it.
>>55713958
what stats before?
>>55714005
2000 to 4000. I haven't noticed any lag ever what so ever, tho.
>>55713958
Seems okay, my non-1000 series hangs around 80-100 but it had a spike to 9000 so that's kind of worrying.
>>55713970
>using the smiley with a carat nose
>>55714012
I just checked mine (also have a 1070) and it reads about 1000 avg. And as you have said, I also don't experience lag of any kind I know of.
What does a high DPC Latency mean?
>>55714065
From google:
DPC stands for Deferred Procedure Call. In its simplest form it is the part of your Windows system that handles driver efficiency. If there is a driver that is taking longer than normal to process, it may prevent other drivers from being processed in time.
Watch this video: https://www.youtube.com/watch?v=uxnFv8cnjZA
Maybe I haven't experienced this issue because I have a G-Sync monitor and it covers the stuttering issue?
>>55713958
>>55714012
>>55714065
I didn't even realize how bad that dpc issue is, what the heck....
Optimal latency is 50 and below constantly
I'd recommend you download Latency Mon and take it from there
BIOS upgrade, upgrading your PC’s hardware drivers, including the video, motherboard chipset and any other hardware
then comes display latency
>>55714107
I have a 120hz monitor, is that part of the problem? I heard people have this exact problem with higher refresh rates
>>55714121
Higher framerate just mean that the time betwen frames are sent to the monitor is shorter. Thus, any deviation in that time frame will be more noticeable.
>>55714121
No it isn't but you got to take it bottom up so it all starts with the bios (and a good psu)
make sure you have a stable overclock or your voltages are constant - using adaptive
disable c states and speedstep in the bios
then make sure your chipset drivers are updated, disable any controllers, make sure USB3 drivers are installed, disable other peripherals... There are many root causes
even not having a mouse driver would increase it, high dpc can be translated in some time microseconds that disconnect the real-time between you and the actual command, it's especially important in audio applications but all around high dpc is terrible and can lead to micro stutter, sometimes you wouldn't even realize it
>>55709911
More like full of pajeets
>>55714197
That's AMD
>>55713807
amd was mostly fine until crimson
I have no idea what they were thinking other than "we gotta appeal to the 12 year old demographic"
>>55714133
I watched that exact video before, and I noticed that even with a lower framerate, the fury x looked smoother on several occasions. Looks like to me that DPC is a big problem on my pc.
>>55714116
>pic related
I've run LatencyMon some minutes ago and Hard Pagefaults max. value was at over 27000.
I've looked into the drivers tab and found out that the nvidia kernel has the highest execution time (3 ms), I will install the new driver and test again
>>55714219
Yeah well use DDU and don't fret, it's a hardware issue so your point of reference is other users with the same config.
>>55713807
The issues started on 364 and that's when vulkan came in and nvidia probably tried implementing async and with windows 10 it botched up? Really not sure but instead of rolling back and rebuilding it just got pushed forward
>>55714275
ok, so the new nvidia driver fixed the latency on their side, but I still get a shit ton of pagedefaults and also have the dpc still hovering around 1000
>>55713958
DPC latency checker is inaccurate on W8/10. Use latencymon.
>>55714365
pagefaults are inevitable
if you want to resolve the problem you need to optimize everything
search cs go tweaks 2015, it's a steam community page - not that I play cs go but it's a great guide
and black viper to tweak your os
be careful with you download and if you're running some botnet good luck with that]
another possible cause is that drivers need to be loaded before the win10 install - usb3, intel sata controllers... so you're maybe running on defaults. download them on your mobo website, gl
>>55714107
Holy shit the 1070 looks terrible wtf is happening?
>>55714219
Step it up senpai ;^)
>>55709961
AMD hasn't even fixed the power draw issue. They reduced it but it still draws more on the 12V from the PCI-e slot than it should, it's just that people look at the 75W max and think it is okay when that 75W is split between 12V and 3.3V. The 480 is drawing around 75W from the slot but less than the maximum 3.3V, so it is still overdrawing 12V.
They fucked up by just not putting an 8 pin on the reference model, and they can't fix that with drivers.
>>55714208
mine looks the exact same
>>55713970
They should fix the problem of being 90% slower than Nvidia cards on AMDs cards in 98% of games
>>55710047
But but you get more fps and the privilege of paying nvidia tax. AMD will never be able to compete with that.
>>55708645
Make sure to disable wifi, bluetooth, and anything wireless before checking DPC latency. Those things interrupt the computer like crazy so disable those before checking to see if your GPU is the culprit in the first place.
>>55708645
hahahaha enjoy laggy piece of shit.
it's a feature.
Just tested SFV, a game I was having unexplained stutter on during matches. Seems to have fixed that - need to test out some other games/scenarios.
>>55713176
Who cares. He's happy.
>>55714203
50% of nvidia engineers are indian.
dont believe me? go to the nvidia campus, it's open campus and you can sit around in the lawn
Has the gimping already begun?
>>55710060
she's an inspiration to pedophiles everywhere
Should I be worried? It never goes under 1000, despite not having any issues... drivers are all up to date.
>>55717388
Go to just about ANY software/tech company right now and it's like this.
>>55718184
>test interval 1000µs
>>55718489
That's the refresh interval actually. It doesn't allow you to change it anywhere and it should still measure lower.
Excuse me while I enjoy my slightly slower, slightly cheaper, more power hungry poo in loo GPU with low latency and no audio and video stutter (I run a DJ stream with ASIO driver so low latency is a must).
Suck it up nvidiots!
>>55713958
>>55714116
I'm not sure I'd rely on this utility for DPC on Windows 8 or 10.
They even have this disclaimer on their website:
>Windows 8 Compatibility: The DPC latency utility runs on Windows 8 but does not show correct values. The output suggests that the Windows 8 kernel performs badly and introduces a constant latency of one millisecond, which is not the case in practice. DPCs in the Windows 8 kernel behave identical to Windows 7. The utility produces incorrect results because the implementation of kernel timers has changed in Windows 8, which causes a side effect with the measuring algorithm used by the utility. Thesycon is working on a new version of the DPC latency utility and will make it available on this site as soon as it is finished.
I can cause the latency to permanently spike at 1000 microseconds just by moving my mouse around.
When running this software in these posts >>55714219 >>55714208 I'm told that;
>Your system appears to be suitable for handling real-time audio and other tasks without dropouts
Running a 780 for whatever that's worth.
Man I had DPC issues for a day because bad motherboard drivers and it drove me fucking insane, I honestly feel bad that nvidia is so bad at fixing this shit
>>55718708
>They even have this disclaimer on their website:
Oh, that would explain this:
>>55718184
>>55710047
>he really thinks anyone here actually buy cards
>>55713958
1070 without the hotfix
>>55710047
I haven't had any issues, though I've barely had the cards for more than a week.
>>55708645
why is this even happening?
>>55714209
Catalyst was rubbish, antiquated software leftover from the ATI period. It was about time AMD relaunched the software.
>>55719750
Maybe Pascal is constantly telling the CPU to submit to its demands. :^)
>>55713958
You do know DPC checker is broken on win 8 and 10 right?
What are some signs that I should be looking for with this issue? I haven't noticed shit for the month or so I've had my 1070.
>>55719970
If you don't have any issues, then don't give a fuck. :^)
>>55720016
The way people in this thread talk it's like the cards are unusable or some shit
>>55710060
the fuck?
what is dpc latency? can someone explain to me?
>>55720220
It is how long the OS is busy processing driver interrupts. Higher DPC means that the drivers hog your cpu more, and it increases the minimum amount of time it takes for the system to respond to *anything*. For low-latency stuff like audio processing, this can be critical.
>amdfags ignoring the past
>>55719970
When I found out about this I checked my DPC latency too and it was fucked. Audio is fine and games are very smooth (at 4K even), so I don't give much of a fuck really. 1080 SLI is an absolutely vast improvement over 290X CF I had previously.
>>55720429
Obviously doesn't mean they shouldn't work on it. But for this to be blown out of the water like it is, when the same people claimed AMDs latency was no big deal is a joke.
>>55720463
They very much should work on it, of course. I want it fixed, obviously.
AMD's problems weren't about DPC latency though, they were about frame time consistency, they had huge issues with that, especially in CF. Hell, they still do in CF in certain games, like TW3, even with HW and HBAO+ disabled.
If you use audio tools that use ASIO etc then high DPC can be the difference between a successful studio session and an absolute disaster. I DJ and use ASIO. When my DPC was fucked I had constant dropouts. So yes it is very important for a lot of people.
>tfw been using my 1070 for a month and did not even notice this
>>55720381
>nvidiots ignoring the past
>>55720758
>>55720741
How do you not, are you deaf or something?
>>55720741
>install hotfix
>almost nothing changes
ayy
>>55720834
Because I have no idea what DPC latency is, something about videos and audio being suttering I guess.
>>55720879
saw someone post a before after where theirs was just as bad as that, then dropped to 30-50. So it works for a lot of people, they'll have to keep working on it I guess.
>>55720954
It's not very significant, it did drop to slightly over 1k but I'm getting some random spikes up to 4k.
I dont see any difference in video quality or audio tho.
>>55710154
Why would you need a vulkan or dx12 fix when the 1060 is far faster than the rx480 in system that use a mid range CPU that most users have.
And seeing as the 1070 and 1080 don't have any real opponents, i see no reason why nvidia need to push for better API integrity with their drivers right now.
>>55721219
Because all consoles use dx12 hardware, meaning that all PC ports will have them in a few years. If NVidia doesn't fix dx12 support, then their hardware will age horribly, while GCN cards keep getting more and more performance boosts.
>people using DPC with Windows 8/10
I hope you fools realize that it will show incorrect numbers, because it's not supported on those systems. Use LatecyMon
>>55714833
>>They fucked up by just not putting an 8 pin on the reference model
You have no fucking idea how this even works
>>55721535
Better than you.
I seriously have no idea what the fuck this shit is.
Is it actually fixed?
That last line can't be good after the update.