I didn't think we had an official SLI profile from NVIDIA yet, do we?

Yes you are right, no we don't have official support, but I was wondering if anyone has tried. I know some people are using sli and have seen a difference. Just wondered if anyone has tried the newer cards with 4k and sli.

Curious if anyone has tried a 4k g-sync monitor with P3D...

There has been discussion on Nvidia forums that it is difficult to get full 4:4:4 color at 60Hz video with Samsung and other 4KHDTVs. You would notice this with simple text screens. Has anyone found a combination that works OK so that the 4k tv can be used as a monitor?

I didn't think we had an official SLI profile from NVIDIA yet, do we?

I'd love to learn what LM has done to garner support from NVIDIA on this--seems odd to me that this hasn't happened already considering the size of LM as a company even though market share for P3D is low.

I'd love to learn what LM has done to garner support from NVIDIA on this--seems odd to me that this hasn't happened already considering the size of LM as a company even though market share for P3D is low.

 

Doesn't seem odd at all to me.  

 

Has been discussed before, if LM purchased large volumes of nVidia graphics cards I'm sure they would have more pull with nVidia, but they don't.  It's all about sales and feedback.  Would nVidia have a reason to want to partner with LM ... don't know?  LM is a huge company with many projects from space exploration projects to earth mapping projects to military projects ... I suspect however the processing power LM use for other parts of the company are SPARC based or even more exotic hardware ... but most likely not nVidia. 

 

What gets nVidia's attention is Sales volume and feedback ... the two go hand in hand.

 

Cheers, Rob.

 

 

OK so that the 4k tv can be used as a monitor?

 

Yes, it's my daily work monitor and it's extremely more efficient to my work flow ... I can have code windows up (VS 2012), SMSS (SQL Database manager) open viewing my data, specs open (word doc multi-page) without ever having to re-arrange windows or re-size anything.  That alone speeds up my development process and most importantly removes potential for mistakes.

 

30 Hz is perfectly fine for 12-14 hours (at least on my Sony it is) ... no headaches, no flickering, all nice and smooth.

 

Cheers, Rob.

Doesn't seem odd at all to me.

Cheers, Rob.

Yes, I already mentioned 'despite the low market share...'  

 

My statement was that I'd love to learn what has been done in this regard.  You say it's been discussed before.  Do you know what LM has done towards this end?  Have they attempted to negotiate a customization by NVIDIA driver developers for P3D?  What would would you imagine it would cost LM $$ wise to pay NVIDIA to optimize the driver for P3D?  Were LM a small company with little to invest and a very unknown product future I could understand this, but again LM is big, they seem to want to take P3D into the future, and since they probably spend significant resources addressing things affecting performance it seems this is a very important area to exploit in light of the fact many to most users will be using NV GPUs--especially w/ regard to SLI as you've pointed out in the past--much simpler to SLI 2x-4x versus upgrade the whole box.   If it's simply a matter of a very large sum of $$ to pay for driver optimization this is easy to follow--have you heard this was the issue?  I would imagine the combined salaries of the LM programmers must be very substantial annually.  If LM coughed up $50K to make the driver sing for SLI, it could be very meaningful I would think.  I would imagine you could do some serious optimization as an experienced driver developer for NV w/ that sort of time.   4-6 months of full time effort?  Seems doable.

It would have to be a lot more than $50,000 to get nVidia's devoted attention.  Money always talks so I guess in theory a substantial sum would get attention but $50,000 isn't going to hack it (not even close).  Large sums of money are never simple matters.

 

Like I said before, DX12 would be the more effective approach, removes nVidia from the equation for the most part and might even remove a need for SLi.

 

No I don't have specific conversations details or email that have gone back and forth between LM and nVidia ... I doubt that information would ever be shared and rightfully so.  Beau has posted publicly they've been in conversations with nVidia.

 

Cheers, Rob.

If $50K isn't in the ball park which intuitively seems reasonable to me then I've apparently underestimated it.  My fantasy is the driver developer who is very familiar w/ the architecture and controls should be able to optimize performance & capabilities w/ say a half a year's worth of consulting time.  Heck, hire a programmer to do it on the side.  I am truly not familiar w/ what's really involved, so I base the intuition on purchasing custom programming work for our health record database.  For $1000 we've gotten some awesome routines developed.   Tweaking a driver intuitively doesn't seem like rocket science, more trial an error.  But as I say, I don't really have the knowledge about what's involved.