the needs of the many outweigh the needs of the few, or the one.and leave me without a computer ?
the needs of the many outweigh the needs of the few, or the one.and leave me without a computer ?
Clanger when the walls fellthe needs of the many outweigh the needs of the few, or the one.
I'm not sure where the cutoff is in LatencyMon, maybe 1,000? Anything below that it will say it's "Fine", but this isn't always accurate, like in this Nvidia case here. I think they just had to draw a line somewhere for guidance, but if it's spiking every 10 seconds up to say 500 or whatever below the cutoff, it will manifest as problems, such as audio dropouts, lower 1% framerates, stutters, etcetera.i did a latmon and it had spikes with dx and nvlddmkm.sys but lasso said that it was perfectly fine, no issues
that was empire strikes back wasn't it ?the needs of the many outweigh the needs of the few, or the one.
I'm not sure where the cutoff is in LatencyMon, maybe 1,000? Anything below that it will say it's "Fine", but this isn't always accurate, like in this Nvidia case here. I think they just had to draw a line somewhere for guidance, but if it's spiking every 10 seconds up to say 500 or whatever below the cutoff, it will manifest as problems, such as audio dropouts, lower 1% framerates, stutters, etcetera.
Yeah, the other thing that makes this difficult to resolve too is not everyone is sensitive to this stuff, even when it comes to framerates and not DPC. Necrosaro commented about this in the past, and funny enough I have the same situation. I play Nintendo Switch with my wife, and that console is so underpowered for most of the games it runs, we'll be playing Overcooked for example and the framerate is already capped at 30 to begin with which I can barely bring myself to play on, but it will constantly dip down into the 20's and it's so noticeable to me, to the point where I just couldn't play the game anymore.no audio dropouts, and averaged around 400+ let it run for a good 15 mins. did have to shut down AV and a few other background things, but i hear ya. it didn't hit peaks every few seconds, it mostly just picked up other things after initial nvd and dx spikes
Yeah, the other thing that makes this difficult to resolve too is not everyone is sensitive to this stuff, even when it comes to framerates and not DPC. Necrosaro commented about this in the past, and funny enough I have the same situation. I play Nintendo Switch with my wife, and that console is so underpowered for most of the games it runs, we'll be playing Overcooked for example and the framerate is already capped at 30 to begin with which I can barely bring myself to play on, but it will constantly dip down into the 20's and it's so noticeable to me, to the point where I just couldn't play the game anymore.
My wife on the other hand is like "You're crazy, I don't see any problems." Then I go to dash around the kitchen and my character falls through the map and dissapears, and I'm like "Uh huh. You were saying?" lol. Unrelated bug obviously, but summarizes my experiences with gaming over the years
i'll run a latmon ( latencyMon home edition v 6.71 - resplendence.com )
see what that 1 says........
additional note. i have steam running and a few edge windows open, while the software is running
This is actually good stuff that you posted Taosd, thank you for the txt log. So I have a few quick takeaways from your report, which shares some similarities to Necrosaro's:
Your Nvidia driver isn't spiking very high at all, only up to 188 microseconds under a load (you had things like Twitch running, etcetera) which is actually quite good in this circumstance. It may be better to run the test again though, for just 5 minutes, and without anything else running, on a fresh reboot of the PC, that way we can more directly compare your "DPC Count" numbers to Necrosaro's results and see what that tells us. It's kind of hard to know if those DPC counts are good or bad when things are running like Twitch for example.
It's possible though, from your Nvidia number that maybe you too have some tweak on your system that is preventing the huge spikes so many of us are seeing. Some of the DPC Counts in your log are pretty high, but most of that is likely just coming from the network adapter being used, plus also McAfee is assuredly causing the ndis.sys and tcpip.sys to be higher than they normally would too, so I suspect that's the issue there.
If you don't mind posting, what are your PC specs outlined in this post:
https://www.ntlite.com/community/in...iver-nvlddmkm-sys-dpc-latency.2924/post-30597
Are you currently using any NTLite xml and/or things like registry tweaks and such on your current system?
Mine is a much older one, had a 8120 originally then took my friends 8350 when he upgraded just as a heads upHrm, interesting that Taosd and Necrosaro have the same Nvidia graphics (1070) as well as both having AMD processor.
In this thread, Slender also reported the combo of AMD processor with Nvidia graphics and had no latency issues (though he was complaining of mouse jitter), while Savitarax has an Intel processor with lots of cores and had DPC issues until he manually assigned the Nvidia driver to another core.
This all plays into the "What we know" stuff I talked about in my recent summary post, where I think what's happening is people that have a lot of cores, tend to have a good chance that Windows will out of dumb luck move sensitive drivers onto different cores which is why some people don't have the issue.
I'm not well versed in CPU affinity though, am I wrong in assuming how that works? Do the drivers themselves get coded to determine what core to move to, or is it fully up to Windows to automate, or does everything just get thrown onto Core 0 and then the other cores get used in other ways?