Super Vr-4

[Carlos Beirise]

Im thinking to buy a Quest 3 and a 4070 super. Would I be able to run ultra with this combo? Would my 12100 be enough for a 4070 ultra? I've looked at some of these "bottleneck calculators" and have come to the conclusion they are full of crap because other sites using actual benchmarks in games have proven them to be inaccurate.

I looked at the 4080 but it appears to be twice the price with only 30% more performance. I can get a 4070 super for $589 US and it looks like the 4080 would cost me twice that but wouldn't double my performance. The prices on the 4090 (at MSRP if you can find it) are 3 times that of the 4070S but only are 70% faster so I don't see that as such a great deal either. It's also out of my budget.

If anybody has a quest 3 could you please give me an idea of what sort of performance you get with a card other then a 4080 or 4090? I'm wondering if I'd be better off trying to run the Quest 3 on the 1080ti or if I'd see a big improvement in the quest 2 with the 4070S.

Historically, Nvidia has followed a poorly-selling generation up with a well-priced gen. The 1000-series was well-priced after they set the prices of the 900 cards too high. The 3000-series started off well-priced after the poorly selling 2000-series (until the mining boom happened). So I think that there is a decent chance that we'll see a good generation at the end of the year, also because AMD messed up this gen and should hopefully fix that somewhat for next gen, putting pressure on Nvidia.

I would think that the Quest 3 is the less risky buy right now, as I don't see a better price/performance headset entering the market at the mid-end for the next 2 years or so. The better optics and higher resolution screen should improve the experience even with the 1080 TI. So I would suggest getting that one first, enjoying it for a year and then upgrading the GPU in early 2025.

When it comes to VR, comparing graphics card price to performance does not work very well. If you are only playing on a monitor, then those kinds of calculations can make sense since most new graphic cards will provide at least adequate performance. VR headsets are so much more demanding than monitors, and there is a point where even a brand new graphics card (say an RTX 4060) will not provide satisfactory performance. I have an RTX 3080 with an HP Reverb headset, and I feel like I am now at the bottom end of acceptable performance (especially since my FPS has dropped by 30% for no discernible reason this month). The RTX 4070 is supposed to be very close to the RTX 3080. If I was thinking of buying new today, I would definitely not go below an RTX 4080, and would likely hold my nose and look at the 4090. But Aspje's suggestion of waiting for the next generation of NVIDIA chips makes a lot of sense to me; lets just hope that there is a next generation, since NVIDIA's attention seems to have shifted to the AI business.

The difference between successive generations of video cards is limited, so more expensive cards of earlier generations typically beat newer generations. Thus a newer generation card is not necessarily sufficient (or preferable over an older card), especially if it is a lower end card like the 4060. In fact, the 4060 is only a tiny bit faster than the 3060, perhaps the worst generational uplifts ever. If you look at the specs of the chip, the 4060 is more like what the 4050 should have been.

I personally was ruined by trying out VR with IL-2, but a year ago I want at least a 3080-level of performance and preferably 3080 Ti or above, at these kinds of prices. But now it has taken too long for me to still spend this amount of money for a card that will be replaced in a year with a new generation of cards (that have a good chance of being substantially better value). Also, I feel morally 'dirty' if I reward Nvidia for what they did.

But of course everyone should make their own choice. The 4070 Super does seem to be the best value for a card that can actually do pretty well at VR, but it's still not amazing performance or performance/$ and the 12 GB hampers its longevity. Because we are now relatively close to a new generation, the value of the card as a current-gen card is reduced, also because Nvidia has now been introducing a lot of features that only work on the newest generation. So waiting a year potentially also means that you get a new feature. And as I said, I see it as likely that the prices/performance will improve more for the next generation.

This is a recurring theme. If you believe you "must" buy now on a budget you set that is your choice. As has been stated above though, there is no certainty that VR performance in GB will be consistently maintained in future updates. If you spend for gpu hardware that is VR-marginal, that is a risk you assume. Best of luck and please report back if you buy!

I think that you are confused about how headsets work. The screens in the Quest are angled and the optics cause distortion, so if the software would just show you the image of the game like on a flat screen, it would be awful. So they do all kinds of corrections to make the image look right.

These corrections cost resolution, so to actually render at the Quest 2's combined resolution of 3664x1920, you actually need to render at 5408x2736. That's about double the pixels (and also double the pixels of a 4k screen), which is why VR is so demanding if you want to max it out.

So by default the Quest 2 is not actually rendering the maximum quality, because it wouldn't work for anyone who doesn't own a 4090 (and not necessarily even then). Your 1080 Ti certainly is not rendering maximum quality on the Quest 2. (Note that you can change the render resolution yourself and this is one of the things you can tune to get a better image, rather than change the ingame-settings.)

So you may ask whether it makes sense to upgrade to the Quest 3 if you are not maxing out the Quest 2. The answer is yes. First of all, the optics are better on the Quest 3, so that will make things look better on the same resolution. Secondly, the Quest 3 seems to have fewer distortions, so it probably 'loses' fewer pixels when doing the corrections. Although unfortunately, Meta have not told us that GPU render resolution is needed to actually render the full resolution on the Quest 3 displays like they did for the Quest 2. I expect that it is lower than for the Quest 2.

I suspect the mid range 5000 series will not be out for a year and I don't plan to spend $1,800 US or whatever it ends up being for a 5090 so I'm thinking I might pull the trigger now on a 4070 Super as I think it would be a big improvement over my current 1080ti.

I might be it , but nobody do benchmarks at 4k if do want benchmark CPU so I could not make all out from 4070 with my I-9 10900 because CPU is bottlenecking. At 4k and above with the best CPU, GPU starts bottlenecking. 4070 ti is perfect for 2560p or even 3440p. But for VR I would recommend at least 480. Anyway good CPU is also a must have , like 7800x3d or i9 13900.

It certainly is an overwhelming landscape trying to get satisfactory VR performance with all the GPU and CPU choices out there.I have a Quest 3, an I7 OC at 4.9 GHZ and 48GB DDR4..... and a GTX 1660.I get a steady 65 FPS but its just about unplayable due to the shimmering and blurriness... Tried ALOT of tweaking but a GPU upgrade is needed,and like the OP, I am looking at a 4070 too. I would love to hold on until we get some better options($) this year,and maybe someone here has some ideas on how to get my rig at least usable.

I've a Quest 3 and a 1080ti. Using Open composite. I only play SP. Everything is clear, readable and enjoyable. 70 to 75 FPS. I'll upgrade when my system blows up. Until then, I'd have to experience someone else's system first hand to see what's possible with new components, but it'd have to be life-changing for me to spend the sort of money these things cost. Somehow, I don't believe it'd be that different.....

3a8082e126