Hello,
I backed the 8K headset and hopefully it will be delivered in the next 2-3 weeks but I still need a new graphics card.
My favorites are either a GTX 1080ti or a RTX 2070 Super.
Which would performe better?
The 1080ti has 11GB GDDR5 RAM, the 2070 only 8GB but in GDDR6,
Would the RAM make a difference?
What do you think is the better buy decision? 
0 voters
While I own a 1080ti; I would Reccommend the Super 2070 (formerly 2080?) Due to newer features the RTX brings like FFR.
The extra 3gigs of memory the 1080ti may givre an edge on some games. But if both cards are similarly priced the newer card likely the better grab. What is the difference in price?
Also found this one.
The price difference is about 20€, while the 1080ti is the cheaper one. 
Yes for the new features it would make sense to choose the newer one. But if the 1080ti gets higher settings in VR games I would choose this card.
Also I have a Intel 4770K CPU and 16GB (8GBx2) DDR3 RAM, do I also have to upgrade these components for VR?
This is my first VR headset so I have no idea how hardware hungry it is…
I would choose 1080ti because its best price&value&performance and i didn’t heard any good feedback about rtx or ffr in vr. Correct me if I’m wrong, i saw ffr discussions, but can’t remember people praising it.
p.s. vr can be very demanding on specific games, even 2080ti can struggle in some scenarios. Cpu and memory can improve it, but i wouldn’t necesarrily change them before you start experiencing iissues
I’m testing the RTX Super cards features with the Pimax 5k+ recently and have to say that I’m not impressed by the results of current Pimax’s FFR, visually it works in a very basic and not pleasant form, tending to generate an annoying effect for the eyes, and the performance gains are still small.
Performance wise, the RTX 2070 Super generally tends to be just a little slightly faster than a 1080Ti but the results may vary with some titles that still show better performance on the older card.
The price/performance ratio is obviously in favor of the 1080Ti especially if you’re going to pick one in the used market where these are abundantly available now for very reasonable prices, but RTX Super cards are probably a longer lasting investment.
Go for the 2070 Super. You might lose a bit of raw performance, but you’ll gain RTX exclusive features like variable rate shading, which is important for foveated rendering.
Yeah, the aliasing from FFR worries me a bit. I wonder if it’ll be noticeable in eye tracked foveated rendering. Personally, I haven’t tried FFR. I wonder if Pimax is trying to add blur on top of the lower quality areas to reduce aliasing.
Would be better
I would get the 2070 Super. I try to buy something “current” so that I get a longer useful life out of my video card.
The raytracing is quite cool, but rarely used in games (yet). If you do go for the 2070, be sure to try the recent RTX port of Quake II, available on Steam. It breathes new life into a game I played years ago (when Q2 was first released).
The demo version (just the first 2-3 levels) is free and the full game is cheap. It’s an interesting mix of new technology and retro gameplay. I think it should be playable in VR using VorpX, but I haven’t actually tried that.
Yes newer one is always better if in doubt.
It’s still in a very raw form, the resolution scaling and difference is way too excessive, leading to a very blocky pixel effect that is noticeable even out of the corner of the eyes, this is not how a foveated rendering is supposed to be.
The resolution scale and difference should be less blatant, otherwise is just a way to worsen the perceived image, and the improvement of the definition in the center is still absent.
Well, hopefully Pimax can improve on things. A blur would be very good for this, but I’m not sure if Pimax can force something like that onto a game. Although I suppose the blur doesn’t necessarily have to be entwined with the game’s engine, so perhaps it can simply be overlayed onto the screens somehow. I’m not a game dev though, so this is just my two cents. Maybe they could make a separate program unrelated to the game that just adds blur on top of the game.
I think that a blur effect would be very helpful for FFR. It’s possible that a blur “post-processing” step could be done by nVidia, Pimax, and/or the game developer. That is, I think any of them could add a blur to the outer, lower res regions.
As it is, I find FFR (even on the lowest setting) to be too distracting due to aliasing, so I leave it off.
Oh I’m sure it would. I’d reckon it’d do a ton to fix the pixely look of FFR.
One thing to note, adding a blur might not be feasible in VR (at least on some systems). Any post-processing is bound to increase lag (video frame delay). Too much lag and users get VR nausea.
Yeah, but blur seems to be a pretty simple type of post processing. I could see why something like Anti-aliasing might come with a lot of lag, but simply blurring the image might not have nearly as much latency.
Actually, blur would probably incur a significant slowdown. (I have a background which includes writing image-processing algorithms.)
A simple post-processing effect would be something like color-tint or increasing contrast. Each pixel is sampled once, processed, and output.
A blur filter must examine the surrounding pixels, do an arithmetic mean operation over at least 5 pixels, including a divide operation, to output a pixel value, That’s 5 times the memory accesses (although caching values would reduce that overhead somewhat).
There would definitely be a delay, the only question is whether or not it’s a noticeable delay.
I believe foveated rendering can have a latency of up to 80ms and still be fine iirc. 80ms is a pretty long time in the realm of latency so I think it’d be okay?
Okay after doing a bit of research I think 70ms is the maximum for foveated rendering latency. https://research.nvidia.com/publication/2017-09_Latency-Requirements-for
Although, I suppose in reality it doesn’t matter as much because we can just increase the high resolution areas if the latency is too high. That should give more leeway for the blur to catch up.