VR experiences IL-2 Korea

[Dash,Polder]

10 hours ago, Voxman said:

Finally got around to this, but, this still alerts Windows Defender that there is virus: 

 

Anybody else getting this message from Window when extracting the file?

Because it's never been submitted to Microslop for digital signatures, and never will.  All files not submitted nor approved by big bro will trigger the alarm.  You can't reasonably expect these small programmers to shell out to the corporation for a simple app, that's called extortion. 

The solution is give your game folders a security exemption and unzip it somewhere in one of them.  I install all my games plus the utilities I use in them in an exempt folder, there is no valid reason on earth windows needs to be actively snooping every file as it's drawn forth and used while your playing.  Try it some time, your over flights of major cities will smooth out big time.

[kmo1619]

I like technology and I tend to use it to solve problems. So instead of blindly trying different settings, I started collecting some actual data.

With the help of AI, I created several analysis scripts and basically recorded a huge amount of system activity while playing, then had the data analyzed with AI. Just to be clear: this is going to take a while, because we are talking about ENORMOUS amounts of data. Files larger than 10 GB cannot even remotely be uploaded and processed in a single pass.

But even at this stage, the results are pointing very clearly toward the CPU. More specifically, mission.dll seems to be a major factor. I assume it not only handles the mission itself and its triggers, but probably also a large part of the AI logic. At the moment, it is consuming noticeably more CPU time than the game engine itself. Once the data is fully processed, I may be able to put together a table showing the differences.

Since the AI was improved significantly in the latest patch, my current assumption is that mission.dll has become more complex and more demanding as well. That would also help explain why performance has dropped despite the reduced graphics quality.

I tested several graphics settings, and their overall influence was relatively small. GPU load dropped noticeably with CheekyFoveatedDLSS, and CPU load followed to some extent, although much less dramatically. In terms of actual performance, however, there was very little difference for me between CheekyFoveatedDLSS and MSAA 2x. DLSS gave me more GPU headroom, but the micro-stutter appeared in both cases almost exactly the same.

I also noticed that HAGS had been disabled on my system. Enabling it improves the situation slightly, but it is definitely not a game changer.

So far, the analysis suggests that the main issue may be the way the CPU workload is distributed. Several cores are being used, but more than 80% of the relevant load appears to be concentrated on a single core.

I’m going to keep digging into it. My next step is to analyze the CPU load as precisely as possible and try to identify exactly where the bottleneck occurs.

 

Edit: running in 9800x3d ,5080, 64RAM , QUEST 3 and VD USB 

Edited yesterday at 03:13 PM by kmo1619

[Aapje]

On 10/9/2026 at 8:42 PM, YoYo said:

I really need to give VD a try, but I'm waiting for it to come out of beta for Link (is there any word on when that might happen?)

The issue is that Ethernet over USB now only works in VD because Meta added support for it, but only in an unofficial way that is not guaranteed to remain working. VS can't really market this as a supported feature when the rug can be pulled out from under them at any time by Meta.

 

 

[Aapje]

On 10/10/2026 at 8:06 AM, YoYo said:

In any case, we won't be seeing the next generation of cards anytime soon; we’re likely looking at a two-year wait, plus the time it delivers for actual availability after launch (which can easily be several months or more), not to mention the price, which could end up costing as much as half a car! High-end GPUs will become a luxury reserved not just for the lucky few, but for the wealthy.

Could be, but there are also rumors about 2027, at least for the AMD side. In general, I have the most hope for AMD cards as the price premium for AMD cards is much lower than for Nvidia right now, and there are rumors about RDNA 5 clocking a lot higher and thus potentially offering a big gain (and they may also make a high end card again).

It's also more a question of when and by how much the price increases will reverse, because the current bottleneck is mainly a mismatch between past investments into production capacity and demand, and there is no fundamental limit to increasing that production capacity given enough time. And the AI bubble may also pop.

On 10/10/2026 at 8:06 AM, YoYo said:

CPUs, for that matter, where we're essentially up against a wall.

There are also some interesting rumors for Zen 6 about clock increases and new packaging that greatly reduces between-chiplet latency, a Ryzen weakness. So that may help.

In general, the early 2 nm results seem to indicate that the power efficiency is drastically better, and given that GPU's and CPU's are now often temperature-limited, this can mean significantly better sustained clock speeds, which is important for gaming.

On 10/10/2026 at 8:06 AM, YoYo said:

My point is simply that developers should focus on optimizing for current hardware rather than waiting for the next GPU generation

I think that they should do that anyway, given that production costs have permanently increased compared to the past and as a result, the cheaper products tend to be quite a bit slower compared to the higher end, so it's not like future 60-class GPUs or entry-level CPUs are likely to blow the doors of the current offerings. So if people with a 5090 and 9800X3D are complaining now, then mainstream buyers of next-gen hardware are not going to have a better experience.

Edited yesterday at 10:07 PM by Aapje