I am at a loss of what to do. Maybe it is because I want the 2080 ti but the thought of giving that much money to Nvidia makes me sad, confused and angry…
So I currently have a 1080ti that I bought July 15th. It’s and EVGA so I qualify for the Step up program. Currently I am in queue to step up to the 2080 and it will only cost me $50 for the shipping. I can cancel my step up and go for the 2080ti but it will cost a further $500 CAD, which funny enough put’s me lower than what Canadian retailers are selling for at a total cost of $1450 CAD for the 2080ti.
That’s a tough question. I’d gladly pay $50 to swap out for a 2080. I’d also gladly shell out $500 for a 2080 Ti (I think it’s absolutely worth having, but $1200 is outrageous). AMD might come out with a better card in 2019, but based on the last couple years it will end up being comparable but not faster than Nvidia cards and there’s a 0% chance that it’ll be $500 or less. So… I guess if your budget can handle $500, then that makes a lot of sense (if not the best bang for your buck [which would be the 2080]). Not sure if that was helpful, but I hope it was.
Somewhat yes thank you. I keep going back and forth because it seems like the biggest difference maker in all 3 of the cards is the memory bandwidth.1080ti and 2080 are similar with the 1080ti edging out the 2080 ever so slightly and the the 2080ti blowing both away in that department.
Seems like the memory bandwidth plays a huge part when rendering at higher resolutions, which is going to be a big factor with the Pimax.
As far as cost consideration goes, it is more so wife agro that is the worry. She had a raised eyebrow when I got the 1080ti.
From what I’ve seen, there’s little point going to 2080, if you have a 1080Ti
There’s a very good chance at this point in time, that I will be putting my 1080Ti on Ebay this weekend, Sunday evening, so that it ends the following Sunday. Then wait till Black Friday to see if I can get a 2080Ti. If not, I’ll just buy one from Scan.
I have a good enough 2nd machine to tide my over till then
No good for VR, but I can play Wurm Online and chat to mates on Discord. That’ll do till the 23rd of November
EDIT: Tomorrow I may change my mind. I keep on changing my mind on this. I’ll get the 2080Ti, I won’t get the 2080Ti yet, I’ll get it, I won’t get it !!!
This is my issue. I keep changing my mind saying “I will need the 2080ti for a good experience with the Pimax” to “Nvidia doesn’t deserve my dollars. I will just tough it out with the 1080ti/2080 until next generation.”
It is a tough call. I was really looking forward to RTX 2080 Ti, but after the benchmarks and price(!) I am no longer convinced it is worth the money. So now I am starting to think about getting some cheap 1080 Ti and wait it out to the next year, when, hopefully AMD will return with 7nm Navi which could compete with the RTX.
On the other hand, upgrading to RTX 2080 for $50 CAD is a different topic. 1080 Ti and 2080 are close enough that I would buy 1080 Ti just for the better price (and 3 GB more RAM), but if you can have either for basically same price and are interested in the RTX stuff, 2080 could be better choice as it will be getting better treatment from Nvidia.
I suppose it does depend on which HMD you are going to get. Maybe if it’s the 5K+ then the 1080Ti or 2080 will be perfectly fine.
I’m getting the 8K X. Those cards will not do.
Then again, I bet it’ll be a year before we get that, which is fine by me. So, could quite easily tough it out till then, then upgrade if need be - to AMD if any better - 2080Ti if proves to be good enough - or wait another year for the next NVIDEA!
Yeah, toughing it out with the 1080Ti or 2080 is a very valid option
I would still suggest to execute certain caution before jumping to 2080 Ti bandwagon. I do not know if you saw some recent videos from @SweViver or @mixedrealityTV who were playing different games.
Regardless if it was DCS or SkyrimProject Cars 2, the games exhibit the similar behavior - GPU was not maxed out (it was sitting around 75% of utilization). It happened only with Pimax, but not with Vive Pro, which suggests that it could be a problem with Pimax’ rendering pipeline that prevents the GPU from running at 100%.
If true however, it could mean that 2080 Ti might not help as much as one would expect. So Better wait until some VR benchmarks with Pimax and 2080Ti are published.
I do not know, it is what fpsVR in @mixedrealityTV videos shows as “GPU Usage”. Here are the examples from MRTV Project Cars 2 session.
I have to however reset myself on my previous statement. In Skyrim he was actually able to run the game at 90 FPS, so having GPU not running at 100% was not wrong. The behavior was observed in PC2 and DCS.
I am wondering if the issue here is what I mentioned above where the memory bandwidth is being saturated by the high resolutions. The 1080ti and 2080 have very similar memory bandwidths 484 GB and 448 GB I think. Then the 2080ti has 616 GB/s.
Since nothing else seems to be maxed this could be it @SweViver s 2080 ti can’t come fast enough.
This could be the reason. It could be however verified even with 1080 Ti by setting PiTool and SteamVR SS factor and FOV to such values that the target rendering resolution for Pimax will correspond to the one Vive Pro uses. If it improves the GPU Usage for Pimax, then memory bandwidth is the bottleneck, if not, something else in the pipeline is.
Well, @VoodooDE just got his two 2080 Ti’s and hopefully @SweViver gets his 2080 Ti in soon . Hopefully between his tests and the TH test on the 2080 you can get a feel for what makes the most sense for you…
To be honest, I have never seen the bandwidth of even a plain 1080, who has 320GB/s , in any game, be it on high res ultra wide 21:9 monitors, or in an headset with high SS numbers, using more than half of that bandwidht…
Some image processing and manipulation sw could go higher than that, but for games…there is almost three times the needed bandwidth in today’s graphic cards, and the higher bandwidth they will push will do almost no difference from a strictly technical point of view, regarding performance numbers.
The problem is another, its the various latencies in both the cores and the memory, that artificially adds bottlenecks and slows everything down, even if the bandwidth could go up to as much as 900GB/s, it will never be fully exploited because of that.
So don’t expect miracles from any RTX card, if Nvidia had used HBM2 on the consumer models, the situation would have been a little better, but anyway not much…the problem will only improve when new types of memory (currently undisclosed because too efficient) will finally come out…hopefully.