HAGS revisited

So I’ve started to see people recommending that HAGS should be turned back on. But on the flip side, I’ve seen many others say that say it’s not a good thing if you’re a VR user.

So, my question is simple. Do you have any personal experience that would tend to validate or invalidate the above suggestions? Did you try turning HAGS on and found it to be a bad thing in VR? Likewise, if you’re a 2D only simmer, did you turn it on and find it to be a good thing?

I welcome any comments and/or suggestions that anybody might have with regard to this question, and look forward to seeing where this conversation goes.

Thanks everyone!

Turn it on for normal monitors. IF you are a VR user and have issues, turn it off.
I dont see any reason to turn it off, unless it causes issues…

I think you might get wildly different answers based on everyone’s config.

I’m 2D only and I find with HAGS off I get slightly worse performance. Mostly on the form of stutters. I don’t chase FPS since I know I can’t sustain more than 40 FPS. I aim for one setting that works everywhere.

1 Like

My experience with HAGS enabled since its arrival on the scene with my games including MSFS 2020 are 1) It provides the same performance as when it’s disabled, 2) It degrades performance.

Since I’ve haven’t encountered a instance where it has provided additional performance I just keep it off. It’s one less thing I have to check when problem shooting or running down a issue.

This is 100% exactly my opinion on this subject. So I keep mine off.

With the latest nvidia driver in VR I still find it to be worse with HAGS on. I’ve also found it to cause some stuttering and stability issues in other VR titles and really haven’t seen any positives from it to outweigh the negatives.

Yes, I have. But first I think the HAGS topic is depending on
1.) If you use VR or not
2.) what graphics card you have
3.) what graphics card driver you use.

For me:
1.) No VR
2.) NVidia 1660 Super
3.) 457.30

I believe that no HAGs gives me less “low fps” moments in Paris. For measurement I used 3D Mark. And 3D Mark produces a little more points with HAGS=off. But the difference is in the noise limit.

1 Like

And for those reading this topic who have no idea what HAGS is it stands for Hardware Accelerated GPU Scheduling.
Ok, maybe you are not much wiser but at least now you know!

I’ve always thought that the correct way to use acronyms was to use the long form first, followed by the acronym in brackets after which it can be used elsewhere but it seems that most posting here assume that we all know every acronym.
I certainly don’t.

11 Likes

If you search the forum for my posts with HAGS keyword you’ll find plenty @KevyKevTPA , but you know these already :wink:

To sum up the various reports from “My VR” and “My 2D” topics:

457.30 VR/2D:
HAGS OFF is better overall.

466.11 2D:
It seems to me more are now finding HAGS ON is better

466.11 VR:
I can’t see any consensus on HAGS

My experience lately is SteamVR 1.17.x (latest beta) with 466.11 + HAGS ON is causing more SteamVR crashes than OFF, and I can’t tell whether HAGS is better or not in VR yet with FS2020 because the game has already lots of other problems in itself, causing too much performance data noise for measuring or evaluating anything at the moment.

1 Like

@CptLucky8, and of course, just to be different, I’m running 465.89. On purpose, even. It’s the first one I found that was as good as, if not better than 457.30, but then when 466.11 came out, it seemed to be a step backwards, so I went back to 465.89.

Heh. Also being different, I’m running 460.89 which, in contrast to newer drivers, has no flickering with G-SYNC on.

1 Like

Best thing anyone can do is try settings like hags on or off on there own pc as what works for one doesn’t always work for another.

I found game mode off and HAGS on works best for me.

1 Like

As mentioned below, lots of different answers to this.
First and foremost is hardware configuration.
Without “HAGS”, Windows scheduling controls every frame delivered to the GPU through a high-priority thread that handles the processing of the data before sending it to the GPU buffer.
Hardware-accelerated GPU scheduling offloads the work from that high-priority CPU thread and instead gives it to “a dedicated GPU-based scheduling processor.”
In general HAGS has been shown to give a 2-3 fps benefit. This is very GPU dependent and some lower end boards that officially support it don’t get much benefit. If the CPU is of the latest generation and the GPU is a geriatric then you are probably better off letting the CPU do the work.

Bottom line…
Turn on your resource monitor and see what the balance is between CPU and GPU. If the CPU is maxed and the GPU is coasting, shifting some workload to the GPU can improve your balance. If you can get the GPU pounding away at 90% +/- and the CPU doing its thing at around 60% +/- then not only are you nicely balanced and not bottlenecking but you have your settings dialed back to maximize your smoothness and quality. It is impossible to get smooth, stutter free graphics with either processor bumping its head on the ceiling.

HAGS on, HAGS off? No way to tell. You will have to test it on your system and see what works best.

1 Like

Same here Game mode off here Hags On with little tweaks in NVIDIA CPL.

@willisxdc, funny that you should bring that up.

I was noticing recently that task manager was reporting my GPU was only running at about 6%! And despite this, I was still getting good performance, in and out of VR, so I figured, well, sheyyt, if I can get that good out of 6%, imagine how good it would be if I can somehow get that up to 100%! We’d be looking at 300-400 fps in VR, meaning that my problems would be solved for all of eternity! (And I was also thinking about all the people complaining that their GPUs were doing something similar, and blaming Asobo for it.)

But I also figured that something didn’t smell right. There was NO way I was getting the visuals I was getting using just 6% of my GPU, so I turned on performance monitoring from Nvidia, which, incidentally, required Geforce Experience to be installed to access. It reported that despite what I had been told by task manager, I was indeed closer to 100% utilization than I was 6%, and I trust what Nvidia’s tool has to say about an Nvidia GPU more than I do Window’s tool. It also just made more sense from a logical perspective.

It also means that all the other people who are complaining that their GPUs aren’t working hard enough are wrong because they’re using Task Manager to judge what their GPU is doing.

I don’t remember what it said I was getting in terms of fps, nor do I care. My eyes are my primary fps counter, and if they tell me something is smooth, then it’s good to go. Yes, if I look down at the ground on the left side of my aircraft when I’m taxiing or on a takeoff roll, I can see some stuttering, but so what? When I’m doing those things, I never look in that direction except purely just to see if the stuttering is happening. In the normal course of events, my attention is elsewhere, which especially in the case of a takeoff roll, is right down the runway, with glances at my instruments, and is exactly where it should be in those circumstances. When looking in that direction, my visuals are as smooth as melted butter, and I couldn’t possibly care less what it looks like when I look in a direction that I never look.

Yeah, I might look that way when I’m cruising (or even still climbing) on autopilot, but when I do that, the altitude makes it look smooth at that point, too.

1 Like

@KevyKevTPA
I generally hit WIN + G to bring up the XBox apps and pin the res monitor then refocus to the sim. I can watch CPU, GPU, RAM, VRAM and FPS all on one small window. Seem to get pretty accurate details.

The problem with the Task Manager Resource Monitor is that it tends to get confused when shifting focus. It was likely showing you the load that the monitor was commanding. You can change what the monitor shows by selecting the process you want to monitor in the ‘Processes’ window. Unreliable and resource heavy on its own.

I have used numerous monitors and seem to get different results depending on which one I use.

I assume your 60% CPU load is the load over all cores. I use the “show load for every core” display of ressource monitor. There I want to have at least one core as close as possible to 100% (the snow white core) and the other cores do their best (the dwarfes, sometimes only 30%). Together they make 60%.

Right now MSFS 2020 version 1.15.8.0 gets 80% to 90% on “snow white”. This is a little improvement.

MS-Windows ressource monitor has a “feature” - I call it bug - that shows you by default the wrong number. You have to select “3D”, not something like “Copy”

Just look for the largest load - this will be the correct GPU load display. - Thanks CptLucky8

1 Like

@PaulFalke IIRC on the 2070S in VR it is Graphics_1 instead (it used to be the last time I checked at least)

1 Like

I assume my 1660 Super is exactly such a “low end” graphics card.