i realize this post was a long time ago but if you are still interested you should know that 400 hours of rendering for any project in max is absurd. that's more than two weeks. that's crazy. i hope you are rendering at least a feature length theatrical film at 4k res. ok i am exaggerating but you get my point. you need to go back and do everything humanly possible to minimize the amount of geometry, the texture resolutions, etc. and i hope for your sake you are not rendering someone else's files. often in this business (esp in architecture and design where they do not have extensive training in cg) you get morons who don't know what they're doing and hand over poorly optimized files to another bloke who then takes the fall for a render that takes forever. i hope you are not in that situation. anyhow, the point is, more often than not optimization is not about software/hardware capabilities so much as it is about the decisions the content creators make in terms of polygon counts, texture sizes, etc.
as for the pro vs consumer cards thing this is something i have been looking at with some interest for a potential future purchase for myself. so, everything that follows is not based on practical experience but on research. take everything i say with a grain of salt. basically from the benchmarks that i have been staring at in confusing for some time it seems that it all depends greatly on the exact combination of gpu with what software in what kind of viewport. for example if you are going to spend a lot of time in cad applications like solidworks, autocad 3d, catia, maybe rhino, etc. (these all handle geometry and viewports in somewhat similar ways) you are going to get a massive boost from a quadro or a firpro compared to a consumer card. these kinds of software are used extensively by a wide variety of professionals in engineering, architecture, manufacturing. basically its a big ass corporate market ready for milking. and they value stable workhorses over powerful ones. even entry level cards like the quadro k600 will run circles around some consumer cards that have much higher specs in these kinds of applications.
on the other hand real time rendering that is accomplished in a similar way to the way videogames are done are going to get boosts from the more powerful consumer cards. from my understanding something like zbrush or mudbox (or editors like cryengine or unreal obviously) work much better with a consumer card with lots of vram than with a pro card. and these days you can find even low end consumer cards with as much as 4gb of vram - having a slow card with lots of vram (or fast vram) means you can open files with bazillions of polygons and maintain a decent framerate, say 20-30 fps in a 3ds max viewport. if you run out of ram of the vram then your performance will tank like a motherfucker. this does not mean a consumer card with lots of vram will crush a pro card in benchmarks for these kinds of apps but it does mean that at the price point of pro cards you can get like several consumer cards in sli/xfire and get comparable levels of performance to a high end pro card. the key issue here is that the midrange pro cards tend to really suck at these kinds of apps mainly because they are not very powerful. a $500 pro card is basically mid to lo range. in the consumer world that's a high end card and you will get better performance in apps that use rendering methods similar to games with a consumer card than you will with a pro card.
furthermore to add to the confusion different applications allow different kinds of rendering in viewports, some using opengl, others directx, etc. and therefore are biased more towards the manufacturer that focuses on certain technologies. for example nvidia does its proprietary cuda thing which has been around long enough now that many applications support it and get massive boosts from that alone. so a comparable amd card in that price range will run slideshows of certain apps that are optimized for cuda. on the other hand amd does opencl extremely well and gets boosts from apps that are optimized for that. to add to that confusion some applications allow you to switch on the fly. for example my photoshop used to lag like a motherfucker until i started fiddling with some preferences and after i randomly turned off opengl (this was before i had researched any of this kind of thing) suddenly my photoshop started running beautifully (it could not get any AA to work but whatevs it didn't bother me).
and if that wasn't confusing enough amd has shitty drivers but give much better price performance. you can actually expect to get decent performance out of some low and mid range firepros. meanwhile nvidia has awesome drivers tailored to specific apps. like the driver for your card for this edition of max. but they are stupidly extensive and in the low range they are total ass.
to make a long story short, go for nvidia quadros at the high end if you can afford it. otherwise go for a midrange firepro but be very careful what you are getting it for. it will give you vast improvments in some cases while in others it will run MUCH slower than a comparable (pricewise) consumer card. to give you a sense of how i plan to address this problem myself - i am thinking a setup with a dual boot of windows would work best. one in which i have a midrange amd pro card like the w5000 and the corresponding ptimized drivers installed and another in which i have a consumer card like a 7870 2gb (more power) or a 7850 4gb (more ram). that means that if i want to do work in zbrush (or fight the evil nazi zombie horde) i can do so on my consumer card and if i want to do cad work (rhino, autocad) i can switch to my other boot. i am now researching this to see if this is a good idea.
in closing i should emphasize that i have a relatively superficial understanding of pro cards performance primarily because i am comparing benchmarks other people have done and also because there are astonishingly few benchmarks like this on the internet. the day a consumer card comes out, even if it's a low end one, chances are some review site somewhere on the interwebs got their hands on a pre release copy and started benchmarking it against every other card under the sun. not only that but the benchmarks themselves are things i can relate to. for example i know how metro last light or crysis 3 runs on my machine (like crap) so when i see a mid range card that get 30+ fps in the most demanding situations it's tempting to pull the trigger. on the other hand it's really confusing to look at a mid range pro card kicking the crap out of a high end one in some benchmarks (usually has to do with optimization of gpu design to software or software design to gpu tech) while getting its ass handed to it on others. sometimes a low end pro card kicks the craps out of high end consumer card and other times a mid range consumer card will murder thousands of dollars worth of pro gpu. this makes it all very confusing.
based on what i have seen you can read my analysis above but for yourself i recommend you take a long hard look at the software you use, the kind of optimization you do in your decision making as opposed to relying of powerful hardware, and finally choose a card that is right for your particular needs.
i hope that helps
Facebook
Twitter
Instagram