I never liked SLI, but now with so powerful card and 8GB and small power consumption...
EVGA GTX1080 SC SLI look as excellent choice for mATX RIG...
Example Phanteks EVOLV mATX with nice and small EVGA 850 P2 and off course X99 configuration.
 
They look perfect together and from side...
No way to GP100 beat GTX1080 SC SLI... 
 

 
Later owners could try to OC and if temp become problem they could order EVGA Hybrid Pack and replace coolers.
8GB is more than enough and now video memory can't be problem even for 4K. But I would choose 1440p and clean picture on 27" monitor.
 
 
 
 

Nvidia’s Geforce GTX 1060 Will Rock A 256-bit Bus – GP106 Based, VR Level Performance At An Affordable Price

Read more: http://wccftech.com/nvidia-gtx-1060-256-bit-bus-vr-price/#ixzz49Dq5sONj

add the ASUS tease

http://wccftech.com/nvidia-geforce-gtx-1080-1070-custom-models/ 
 
 
 
and HDR news
 
NVIDIA Pascal Goes Full In with HDR Support for Games and 4K Streaming for Movies

Read more: http://wccftech.com/nvidia-pascal-goes-full-in-with-hdr-support-for-games-and-4k-streaming-for-movies/#ixzz49Rnr41Hj
 

Xavier Zepherious
Breaking news: AIDA64 developers confirm Pascal GP102, GP106, GP107 and GP108
 
http://videocardz.com/60289/breaking-news-aida64-developers-confirm-pascal-gp102-gp106-gp107-and-gp108 
 
Tables listed on Website - noting A and B steppings/revisions - ie refresh
 
gp102 has 11 chip numbers
 
gp104 has 15 chip numbers
 

Thanks for posting this. 

http://videocardz.com/60333/evga-geforce-gtx-1080-acx-3-0-superclocked-and-ftw-pictured 
 
 




 
 
 

ZOTAC GeForce GTX 1080 AMP! Series

 
 
The PCB is fully modified and it hosts two 8-pin power connectors.
http://videocardz.com/60325/zotac-geforce-gtx-1080-amp-and-amp-extreme-pictured 
 
 
 
 

http://videocardz.com/60340/asus-rog-strix-geforce-gtx-1080-pictured 
 
 
 

GALAX GeForce GTX 1070 Gamer Series:

 

 
http://wccftech.com/nvidia-geforce-gtx-1070-custom-models/ 
 
http://wccftech.com/nvidia-geforce-gtx-1070-titan-x-killer/ 
 
 



Zotax 1080 PGF
 
 

 
 
 
 
 

Colorful  iGame-Zone
 

GALAX GeForce GTX 1080 HOF and Gamer Custom Models:



 

Inno3D Chills Pascal GeForce GTX 1080 With iChill X4 / X3 Coolers

also caught this tidbit spotted... 3dcenter.org forums
GP104-200-A1 / GP104 PR DT 
GP104-200-A1 / GP104 PS NB - ASUS 'teaser? 
GP104-200-A1 / GP104-PS-DT / GeForce GTX 1070 
GP104-400-A1 / GP104- PR DT 
GP104-400-A1 / GP104-PS-DT / GeForce GTX 1080 / dev_id 1B80 
GP104-725-A1 / GP104-PS-DT 
GP104-950-KD-A1 / GP104 PR DT 
GP104-975-A1 / GP104-QS-9XX 
GP104-985-A1 / GP104-QS-9XX
GP104-725-A1 / GP104-PS-DT
 
um Chips above 1080?
quadro maybe a 1090
 
 
GP104-725-A1 / GP104-PS-DT       ...1085?
GP104-950-KD-A1 / GP104 PR DT...  1090?
 
im assuming these are quadros
GP104-975-A1 / GP104-QS-9XX 
GP104-985-A1 / GP104-QS-9XX

Inno3D Confirms NVIDIA Pascal Based GeForce GTX 1060 / Ti Graphics Card – Performance Hinted Between GTX 970 and GTX 980 Levels, Sub 150W TDP

Read more: http://wccftech.com/nvidia-geforce-gtx-1060-confirmed/#ixzz4AAkWhiX1

 
http://www.nextpowerup.com/news/28396/new-aorus-x7-pro-model-features-dual-geforce-gtx-1080m-gpus-4k-display-and-better-cooling/
 
uses the word gtx1080M. But, that may indeed just mean a throttled back full 1080 chip. Maybe a good way to sell 1080 chips that won't otherwise pass testing.
 
 
 
 

GP102-400-A1 3584 96ROP 384bit = 3dcenter.org forums
gtx1080ti
 
 
kingpin Overclocks GTX 1080 

https://twitter.com/TEAMEVGA/status/740228112563068929 
 
EVGA overclocking Geforce 1080 GTX to the senseless 2835 MHz
 
http://www.nordichardware.se/nyheter/evga-overklockar-geforce-gtx-1080-till-sanslosa-2-835-mhz.html 

So from what I understand they are only allowing 3 and 4 way in certain apps like benchmarking. Just to keep geforce on the leaderboard. Gee it would be funny to see the leader board of 3dmark be all Amd gpus.

what that means is there is no direct support in drivers for 3+4 way except for those games they have profiles for
the cards will still work - just that you will get no scaling because there no profile or support
 
 
with DX12 came implicit and explicit modes for SLI and the game and app developers wanted this
Nvidia does not support directly any more than 2 implicitly
 
so it's up to game developers to implement 3,4,5,6....+?? way explicitly
beside games required them to implement beyond 2 anyways even with dx11
even if Nvidia gave 4 way support in dx11 - if the game developer did not implement it you got 1 card only or 2 or 3
so what's the diff??
 
it's the game developer responsibility now - nvidia only has to scale 2 way well
 
same goes for AMD... they also have to deal with this
 
you can blame Windows DX12 and game developers for this

no this is a DX12 thing
 
app and game developers wanted more control low level access to improve speed - in doing so they made Implicit and Explicit modes
implicit is only 2 way and is same gpu related
 
explicit is any  # way - gpu independant...mGPU 3,4,5,6,7, and so on
so you can have multiple working on one problem - crunch/fold as well
 
and this is solely on the developer to implement
 
under the old way Developers would be hamstrung by the way SLI or crossfire worked..
it be generic - simplied one gpu maker- and one model type
in order to do that and put the onus on Nvidia on how to implement Sli or in AMD Xfire - they would deliver a generic way(driver side)for all games or apps
not specified how to optimize...because some would do little and some would implement Sli differently
Nvidia would provide the best profiles for each app/game in the drivers - but how do you do that for multiple games multiple developers and multiple ways the developers want to use the gpus in tandom(new algorithms)
or didn't write good code (so Nvidia and AMD would be at the mercy of game developers)
how many games are there - you want to write code to be able to handle all the different ways game developers write code or want to use the GPU...no you offer a standard way
so you got basic Hardware vendor implementation -and the bloat,overhead and restrictions that come with it
 
so what you wound up with is SLI that didn't scale well or not at all(in game) or in some cases where Sli was not implemented(game side) would get some partial benefit
 
under the new regime which is what developers wanted so they can create new ways gpus can do Sli or work with other GPU's... different makes and models from AMD,NVidia,Intel, Etc
so they could implement new strategies and new means at the lower level API
 
sort of leapfrog over the barriers and generic means of doing things
so that can make the games faster, offer better visuals and gameplay
 
beside most games didn't not scale well beyond 2-way because at that level it becomes App or Game developer dependent anyways... not a nvidia or AMD thing - and they are hamstrung by the Dx11 and GPU vendor
 
so the game/app developers asked for the Dx12 change - and the mGPU explicit mode
 
http://www.pcgamer.com/directx-12-will-be-able-to-use-your-integrated-gpu-to-improve-performance/ 
 
http://www.computerbase.de/2016-06/nvidia-sli-pascal-nur-fuer-benchmarks/ 
 
https://linustechtips.com/main/topic/380260-multi-gpu-resource-pooling-dx12-demo/ 
 

Square Enix demo with 4 GPUs in DX12... 8k textures
6-12x more more powerful than dx11
 
http://www.tomshardware.com/news/microsoft-directx12-amd-nvidia,28606.html 
 
 
 
What we're seeing here is that DirectX 12 is capable of aggregating graphics resources, be that compute or memory, in the most efficient way possible. Don't forget, however, that this isn't only beneficial for systems with multiple discrete desktop GPUs. Laptops with dual-graphics solutions, or systems running an APU and a GPU will be able to benefit too. DirectX 12's aggregation will allow GPUs to work together that today would be completely mismatched, possibly making technologies like SLI and CrossFire obsolete in the future.
There is a catch, however. Lots of the optimization work for the spreading of workloads is left to the developers – the game studios. The same went for older APIs, though, and DirectX 12 is intended to be much friendlier. For advanced uses it may be a bit tricky, but according to the source, implementing the SFR should be a relatively simple and painless process for most developers.
 
 
PS:
beside a 1080 is almost as fast as two 980's
so 2-way 1080 blows any previous video card and game Sli profile in Dx11- you don't need much else 
and you still have explicit mode for DX12
 
Basically DX11 is legacy and DX12 is the future
so why would Nvidia want to provide old Dx11 sli support in DX12 when mGPU implementation is vastly changed
 

Xavier Zepherious
from EVGA BIOS
https://www.chiphell.com/thread-1601090-1-1.html 

GP100. 

under mGPU  Explicit mode you don't need a bridge - the PCIE has enough bandwidth for multi GPU support including 8 way SLI
 
what part of any # of gpus and any # of different gpus simultaneously do you not understand?
this is what DX12 has... dx11 does not
 
this is something the game manufacturers asked for and for more low level support
now they are able to add more SLI or mGPU support and multiple new methods and ways to improve support where before they were barred by the generic multi GPU drivers
 
so why go beyond 2-way with basic SLI?  -where you need developers to do the heavy lifting anyways - and they need the flexibility of Explicit mode