Is Titan Xp a meme?
What am I gonna do with 200+ extra CUDA cores?
Isn't it better to get 2x1080Ti's?
>inb4 gaymur lelelele
>not a gaymur
Never dual card.
gaymur lelelele
gaymur lelelele
>>59868330
ikr.
SLI 980Ti here, it's shit.
>>59868330
>never SLI
crossfire has been pretty good in dx12
like insanely good scaling
just look at RE7 and tomb raider benches with crossfired 480s
>>59868479
You mean mgpu is good, crossfire is shit no matter the api
If you already spend more then $700 for gayman, might just spend more to get the best
>>59868494
we sure as fuck arent seeing near 100% scaling for SLI m8
>>59868479
The problem is AyyyyMD is a meme of a company for GPUs.
>>59868312
For actual CUDA work dual 1080Tis will be a much better investment unless you need that miniscule extra ram
>>59868568
I want to get started with 1x1080Ti then get the second one later.
Currently I have SLI 980Ti and combined it's a lot of CUDA cores for me (I use 3d renderer called Octane which uses CUDA cores).
Although I'm mad about the heat the maxwell cards generate. Hence why I want to move to the 1080Ti.
My company pays for it.
>>59868457
>>59868479
>>59868494
>>59868560
AFR is a complete clusterfuck in this age of post-processing renderers that use previous frame data for shit like TXAA and screen-space reflections.
AMD was formerly rumored to be focusing on "scalability" for Vega's successor Navi, but even that's not exactly certain anymore.