Hardware PhysX requires one video card to display the graphics as usual and a second card to do the PhysX calculations. Nvidia now requires that both cards be their own even though it's been proven that it will work with an AMD card as the display adapter with an Nvidia card acting as the PhysX processor.What are the dual Nvidia cards need for?
Hardware PhysX requires one video card to display the graphics as usual and a second card to do the PhysX calculations. Nvidia now requires that both cards be their own even though it's been proven that it will work with an AMD card as the display adapter with an Nvidia card acting as the PhysX processor.
http://techreport.com/news/17693/nvidia-disables-hardware-physx-for-geforce-radeon-tandems
I learned something new today.PhysX does not require a dedicated PhysX card. It does improve performance, but many games allow advanced PhysX on a single card. In the case of having a single Nvidia card, you can still use GPU accelerated PhysX. The two card situation is only with an AMD primary card, because PhysX doesn't work on their cards.
I think you got my drift.Technically it's kepler and above 650 ti, so just a guess would be it's a fraction of that maybe 20% of the market? (GTX 660+)
I learned something new today.
I guess that lays to rest the idea of it being locked.
It does not make it clear if there is specific hardware required within the video card. It kind of sounds like it may, though it may be something that many AMD cards have as well.
Technically it's kepler and above 650 ti, so just a guess would be it's a fraction of that maybe 20% of the market? (GTX 660+)
So of that 65% market share we keep hearing about, how many a rocking a 650ti kepler or better graphics card?
PS he also said the 780 ti is Nvidia's new high end. Which was a bit of a slip.
I am so sick of a hand full of people saying that.780ti fastest gaming card.
Titan was not as a gaming card on the market.
Fastest NVIDIA gaming card.
Playing with words and ppl listening half of them.
The high speed 2400 fps captures of vsync on (triple buffered) and off showed how bad vsync is when the panel isn't precisely matched. If you want to experience a gsync like experience you need to be running vsync on with the game frame rate matching 60 perfectly.
I wouldn't go that far. There is still a measurable variance far in excess of a single card in a lot of cases. Sometimes its out done by the increase in frame rate and sometimes it isn't. But its not universally true that microstutter is just gone, because it isn't. Its mostly down at a level where its unnoticeable on the latest nvidia cards and games.
He also suggested strongly the gsync module does a lot more than just supporting gsync. With 768MB onboard it certainly doesn't need all that RAM purely for controlling the LCD but he wasn't willing to say more about what it could do.
The discussion of draw rate was interesting - he was suggesting its at the panel refresh rate which in the event they had was set at 60hz. So I guess in theory on a 120hz monitor you'll get an 8ms draw rate all the time regardless of the frame rate, that has some benefits as well.
The high speed 2400 fps captures of vsync on (triple buffered) and off showed how bad vsync is when the panel isn't precisely matched. If you want to experience a gsync like experience you need to be running vsync on with the game frame rate matching 60 perfectly.
Single digit %'s, I'd imagine. If all it requires is DP imagine how many more monitors the companies could sell if they allowed all DP out video connections to work. I actually would be quite surprised if the monitor manufacturers would go along with a hardware lockout.
Unless there is something else that needs to be in place for it to even work.
Did he mention triple-buffering ? Not once did I hear him mention it. It would be nice to hear him explain why nVidia doesn't implement a forced triple-buffering option for d3d in their driver settings. Or why they don't encourage more developers to include triple-bufering in their games.
He also claimed that running at higher framerates (or: shorter frame durations) would help reducing motion-blur almost completely. From what I've seen, that isn't true. Running at 120Hz gives slightly less motion-blur than running at 60Hz. But to really reduce it to a minimum, you need the strobelight-function.
Did he mention triple-buffering ? Not once did I hear him mention it. It would be nice to hear him explain why nVidia doesn't implement a forced triple-buffering option for d3d in their driver settings. Or why they don't encourage more developers to include triple-bufering in their games.
He also claimed that running at higher framerates (or: shorter frame durations) would help reducing motion-blur almost completely. From what I've seen, that isn't true. Running at 120Hz gives slightly less motion-blur than running at 60Hz. But to really reduce it to a minimum, you need the strobelight-function.