It seems to promise EyeFinity-like options for creating a single display device out of a set of monitors.
And considering that you can get on a Gigabyte GeForce 980…
3 x DisplayPort
2 x DVI-D
2 x HDMI
…having that flexibility would be appealing.
They claim that it’s a patent pending technology. I presume it switches between linking the outputs in the standard way, and routing them through some canvas-splitting chip on the card which acts like a single display from the perspective of the GPU.
I can’t find any screenshots of the settings dialogue.
Does anyone have any experience with this Flex Display?
well I don’t have experience with Gigabyte Flex but I have experience with the Zotac GTX970 which has a simillar feature, which seemd to be a new standard feature from NVidia for there more high-end Gameing Graphics Cards.
To be honest it was/is not really sophisticateing, the Display configuration was really tricky, somtimes it worked, sometimes not, sometimes it broke up the configuration after rebooting the system, etc. It seems to that this feature needs some more development on the vendor side, and the NVIDIA support was not really supportive.
I heared much better things about eyefiniti, so far.
Mosaic is NVidia’s proprietary display teaming solution available on their Quadro cards only.
Flex Display isn’t the same thing (it’s a Gigabyte proprietary technology available on specific Gigabyte GeForce cards)
I am really interested on the outcome of your testing of the Gigabyte Flex feature. It’s a little bit frustrating that there is no common standard for multi-display configuration, sometimes it seems that we are the only one who needs this feature.
could knowledgeable people list reasons for going fullscreen once instead of n-times?
(+) sharing videos per head was a problem before vlc and player(texture). still is with capture devices, they have to be passed to the cpu first, via sharedmemory.
(-) displays have to be the same resolution
…are eyefinity or flex using 1/n resources compared to exteded desktop ?
Flex Display is perhaps more related to which outputs on the back of the card are activated for the GPU to use (rather than the ability to span multiple outputs as one canvas).
Still have to test here…
@ggml - having 1 renderer means having only 1 set of draw calls. Also it can mean that you don’t replicate interim renderers (e.g. Renderer (TempTarget) is feeding into an object being drawn on 3 renderers on 3 separate devices, then the TempTarget might be being called 3 times. Whilst if it’s going into 1 final renderer then it’s definitely being only called once)