Not interested in anything north of 300 to 400 lol. I do wonder if DX12 will open a wider gap between 980 and Titan 12gb though. The spec difference is larger than the difference in performance observed. The constraints of dx11 might be bottlenecking things to where GPU performance is really held back, and this same thing may hinder 390x.

Would be really interesting to see a 3ghz i7 vs 5ghz OC i7 with the benchmarks to see if it is held back by the CPU single threaded elements of dx11.

I guess 390x and Titan Z / X are only for multi gpu maniacs or those that only want a single card max, as CF 290 should be faster than 390x for cheaper (rumored 700+ price) and SLI 970 is definitely faster than the new Titans (even at 4k from what I've seen). Not that CF or SLI are perfect of course.

yup waiting to build new PC, but this card is overkill and overprice, maybe if most of games that looks like Star Citizen released then this type of card is worth it. For now no games that worth to play.

cheaper to buy two cards and sli them, better performance that way too, as for the extra ram.. yeah, not really an issue for the currently available games, and we really shouldnt be encouraging developers to assume large pools of vram are available because what happens is, games dont look any better, corners are just cut and sloppy ports are released.

GTAIV holds up its hand in shame.

To be honest I worried the same about the large RAM of 8th gen systems, that we would get games with really unoptimized memory footprints.

Not being concious of vram usage makes for lazy, sloppy games.

In putting in the work to make the best graphics you can, fit into the space youve chosen to work with, multiple itterations are made as the map and environments fine tuned, texture data for models is optimized, UV's are packed tightly and efficiently, care is taken over material variety, placement, density and overall appearance, without that set limit to work to, things fall apart quickly.

For example, why optimize the textures and fine tune UVW resolutions and packing to fit 2gb if our development machine can handle 12gb?, and were going to "recommend" at least 6gb cards?, we can save money and time by not putting in the effort to optimize this sort of thing or spend the manpower on doing said work, faster development time = cheaper development, less optimization needed = less manpower required, more profit.

Game quality isn't scaling with hardware performance, because the more powerful "recommended" hardware becomes, the more liberties and corner cutting developers can get away with, that's always been the case.

Flooding the market with high capacity vram cards wont make development teams suddenly start using much higher quality textures and effects, not for a long while, all it will do is allow them to do the same shit at a lower price and get away with it, but the PC fanboys will point and say "hurdur, but it does look betterm look this texture here is 1024x1024 but only 512x512 on console!!!, but the level headed would say, well thats nice but does it really effect much overall, and is it really worth the buy in price of high end cards?

p.s. the titan x costs $305 more if you're from the EU instead of the US.

Agreed. It still astounds me how much devs got out of last gen's 500MB consoles cos they were forced to optimize the hell out of their assets and streaming.

Yeah its quite bad, certain companies like apple and nvidia are shameless. They didnt lower the prices for their products here in europe when the euro was 30% more than the dollar but they increase it now when its 1=1 now