videocardz.com
They will need to do something depending on what AMD reveals. While $999 is quite high for just more ram, I also think they would have to make it a bit more powerful and call it a super to justify a price increase.
Again though, it already has 10GB of bleeding edge GDDR6x in its $700 model, as well as the fancy high density backdrilled PCB. Adding extra GDDR6x on the back side won't change the economics of the PCB much at all. Nvidia is free to charge whatever they want, but upping the price 43% for double the VRAM is going to be met with as much derision as Apple's pricing for memory upgrades.
I don't know, it may well be $1000 (although i doubt it) but automatically this will also mean that at $700 and above there will be no competition from AMD which seems unlikely given the indications that we have: (80CU speculation, the PS5 clocks, AMD's slides stating RDNA2 has also IPC improvements, +50% performance/W improvements and that the last years AMD's advertised performance projections was fair) If Nvidia just wants to retain or increase the profit margins what forced her hand to price the 3080 at $700, 3080 will be around +38% faster than 3070 (2080S was only +18% faster than 2070S, it still had 8GB (now 10GB) and same class type (now GDDR6X). What is forcing Nvidia to have so competitive price in the $700 bracket and at the same time to get away with a $1000 20GB version, it just doesn't add up!🤔That could actually be the marketing game. 3090 partially exists to make the $1000 3080 20GB look good.
Well, I see what you're saying and I agree with many, but like I mentioned in my first post, imo it does not add up, my point was that in order Nvidia to price 3080 10GB at $700 they probably had info about the competition forcing them to price it at that level, even if they were wrong, now it's done, AMD will have to be competitive in price/performance with this card (if they have a card at +$700) there is no good scenario for AMD otherwise..., so I don't see how AMD this time can be competitive at price/performance with 3080 10GB and at the same time Nvidia price the 20GB version (same performance as 3080 10GB) at $1000. I could be wrong of course, (but in this case surely a $1000 3080 20GB does not make sense at all from a price/performance standpoint, surely memory sells, but still) in any case we will know sooner than later.
I don't think 10GB is really enough for the next 3 years at 4k, and would probably wait on a 20GB version at a reasonable cost increase. That is absolutely future me's problem though, and today the 3080 has enough to play at 4k without issue. If it strictly is a memory increase, when review time comes and the $1000 3080 20GB scores the exact same as the current $700 it's going to be terrible price/performance. I think you're right on the positioning, and I'd buy the 10GB version or a 16GB Navi equivalent over a thousand dollar 3080 and keep the $300 to spend buying a 4080 when it rolls around. I don't see how Nvidia can reach that far for 10GB without some other meaningful performance increase.
More vram is definitely useful if you don't upgrade every year. I also use the HP Reverb and some apps and games I use in VR use up a lot. While 10gb may be enough for the majority of those that play at sub 4k resolutions, Im pretty sure it won't be fully enough for the next 2-3 years.
You've got it wrong, please read again my posts. Regarding 3080 10GB, I am saying that it is a very good value at $700 and among other I explain why in my view Nvidia should not charge $1000 for the 20GB version since it will have practically the same performance (and also regarding the potential competition from AMD, etc) But since You mention it, 20GB is important for someone to have if doesn't upgrade too often (just not $1000 important 🤭) For crying out loud, we had in Q2 2016 a 8GB card at $239, in three quarters it would be 5 years since. Also we are at a transition period with the new consoles and memory requirements are expected to rise.
Facebook
Twitter