Discussion in 'Videocards - AMD Radeon Drivers Section' started by Eastcoasthandle, Jul 10, 2019.
Thanks for the feedback. Are you on kb/m or controller?
No comparison against NVidia input lag? I guess they don't like traffic...
I had some interesting results with in Prey that I thought I'd share, seems like AL isn't as "set it and forget it" as you might assume. I'm only on a 4 core cpu (firstname.lastname@example.org) so I was pushing around 95% cpu usage and 99% gpu (100+ fps) and anti-lag really didn't like that. When enabled my gpu usage would drop to around 70% and the frames around 70-80 and have that awful stuttery feeling of a bottlenecking cpu.
So I guess saying, "you need to be gpu constrained" is a nicer way of saying "you need spare cpu overhead"
No, not at all... You "need" to be GPU-constrained to see gain from Anti-Lag. Also, there was a problem that in some DX11 games the performance (and GPU utilization) would drop significantly - but this has been (hopefully) fixed with the latest update. Please check it out and tell me if you still see a problem in Prey: https://www.amd.com/en/support/kb/release-notes/rn-rad-win-19-7-3
did someone actually used gpuview to compare the difference between frames?
Keyboard and Mouse. I could instantly feel the difference by using the hot key and toggling it on and off.
so most people dont measure the difference at software level?
Installed the new drivers and I'm getting similar results in Prey. I found a quiet room with CPU utilization sitting at 72%, enabled anti-lag and utilization shoots up to 90%. GPU fully utilized the entire time.
Well now the results are not similar at all, right? The GPU utilization stays high, and you can check with Ctrl-Alt-L to see the EqFPS improvement. The CPU utilization increases and that’s fine - that’s a side-effect if RAL - it will increase only if CPU is idling some of the time, it will not reduce the CPU performance.
If you know how to use GPUView - you are welcome to take a peek...
Okay I did some more tests and it's not nearly as bad as before but there is a TINY performance hit if you're badly cpu constrained. Example: I went into an area where the CPU was pegged 100% and the GPU around 97%. Turning RAL on dropped GPU utilization by about 1% and I lost about 1-2 fps. So, pretty reasonable and I'm not complaining, the new driver definitely helped. I know, I need a new cpu, but I'm squeezing the most out of what I've got
OSD? more like FPS counter only? ^^
BTW, what is that number supposed to be?
Tried in in DX:MD and ESO and it always showed about 20FPS higher than what MSI Afterburner or Ingame Engine reported.
He already explained that.
It's the amount of fps you would need to have for the same amount of latency you are having with anti lag.
I thought I would give it a few days.
What Are Your Impressions of RAL?
Everything from my end is still copacetic with RAL since the last few driver updates. If performance was improved in selected games then great. But I honestly don't play without it now.
I do notice the difference.
I play PUBG mainly right now and the difference is noticeable.
That's odd. I noticed something...
I removed the default XMP profile and manually setup my ram's timing. Bios has a save profile option so I use that.
I forgot to change ram voltage from auto to it's default voltage. When I did I notice a big uptick in FPS now with RAL still enabled. Aiming in BO4 is a bit more snappier now in 40 degree increments. IE: Strife firing at more then one enemy from a 40 degree range (haven't tried farther yet).
Combo moves execute faster now. And I thought before it was fine. sheesh.
That’s so great to hear, thanks! And - I learned a cool new word... )))
I've been tweking my memory too. Corsair Vengeance LPX 3200MHz CL16 2x8GB Kit
I tightened the latencies and set the memories from 3200MHz to 3333MHz and timings from 16-18-18-36-2T to 15-17-17-38 1T.
Here are the results of the OC. I don't have a previous bench, but, read-write-copy were lower and latency higher.
Did you intend to show results?