Was that in reply to my post? If so, please use the simple Reply button to ensure the quote is added to reduce confusion.So Vega is not competitive. That's why nVidia released an unlocked driver for their Titan. Right.
I do wonder why isn't there a monitor that supports both G-sync & Freesync?
Seems they could wire up a bypass for G-sync if needed.
Was that in reply to my post? If so, please use the simple Reply button to ensure the quote is added to reduce confusion.
AMD may not have the resources either, though. NVIDIA is a massive company that primarily produces discrete video adapters... AMD is not. ATI was
I tend to agree, but, that would be against the Sherman Antitrust Act, they can't tell the manufacturer you can only use G-sync, and can't enable freesync.It is almost a certainty that Nvidia forbids it in their G-Sync contract to manufacturers who buy their module. I have no proof, but why wouldn't one of the high end gaming manufactures want to offer both for a sweet premium and mindshare? I can hardly believe it would be physically impossible, especially with a switch to bypass. The simplest answer is probably the right one: not permitted.
I didn't think it was physically possible to skirt around the PCIe interface as it's physically the barrier between the CPU and the GPU? The infinity fabric is a physical interconnect between components in the CPU, not a mystical ether to communicate over? Or am I confused?Mark my words: AMD will create a driver that when a Threadripper CPU detects a VEGA GPU, it will switch the PHY from PCIe to Infinity Fabric, thus reducing overhead and increasing bandwidth.
I tend to agree, but, that would be against the Sherman Antitrust Act, they can't tell the manufacturer you can only use G-sync, and can't enable freesync.
That is pretty much what Intel was busted for, they attempt to monopolize.
I tend to agree, but, that would be against the Sherman Antitrust Act, they can't tell the manufacturer you can only use G-sync, and can't enable freesync.
That is pretty much what Intel was busted for, they attempt to monopolize.
It would be like the Intel scenario where Nvidia pays the monitor companies to not have both G-sync & freesync on the same unit. It doesn't have to be physical payment either, there are many ways to threaten a company...Well, it's just for the monitors with the module installed. Is that act relevant? Didn't Intel pay manufacturers not to use AMD? How is that similar? The manufacturer can do whatever they want with other monitors, as they do now. And we already know GS has strict timing standards. If it is a legal issue, they merely need to state that any VRR on the monitor must meet these standards to use the monitor, that FS cannot quite achieve.
I didn't think it was physically possible to skirt around the PCIe interface as it's physically the barrier between the CPU and the GPU? The infinity fabric is a physical interconnect between components in the CPU, not a mystical ether to communicate over? Or am I confused?
It would be like the Intel scenario where Nvidia pays the monitor companies to not have both G-sync & freesync on the same unit. It doesn't have to be physical payment either, there are many ways to threaten a company...
We need the monitor guys to chime in on what is going on.
As for who would buy them, if it is a high end monitor anyway, people that want the choice of which GPU would be first in line. When nvidia is better, they can use that card, when AMD is better, they could use their card.
Right now, if you spent $1K on a monitor, you are pretty much locked it to either g-sync or freesync.
So Vega is not competitive. That's why nVidia released an unlocked driver for their Titan. Right.
Freesync may not have royalties, but that doesn't mean it is without cost to implement, even if it is only software.
Also you are going to build FS displays without GS anyway, so what is the real value of FS+GS display?
The only people who will really pay extra for a FS+GS are the people who want GS. Everyone else wants to pay less for FS display, so there is very little value add to have FS+GS.
Other than the opportunity cost of not using freesync, what would the cost be?
I don't believe that's how it works. The fabric is within the core itself but isn't some alternative method of communication between the GPU and the CPU. The physical PCIe interface is still a factor.Check out this link: https://www.nextplatform.com/2017/07/12/heart-amds-epyc-comeback-infinity-fabric/
AMD uses Inifinity Fabric links are configurable for PCIe. Since the AMD VEGA also uses Infinity Fabric, it's possible that it also can be reconfigured from PCIe to Infinity Fabric and get the 20% performance boost.
You think software development and testing is free? It will cost money to add Freesync support to a Gsync monitor.
Infinity Fabric is also used for Multi-Socket CPU links, so it's most definitely not an on die only comm PHY.I don't believe that's how it works. The fabric is within the core itself but isn't some alternative method of communication between the GPU and the CPU. The physical PCIe interface is still a factor.
I've not heard of it being used for multi-socket communication (inter-CCX within a CPU yes) but it's definitely not something communicating with the GPU.Infinity Fabric is also used for Multi-Socket CPU links, so it's most definitely not an on die only comm PHY.
They already do that testing when the same people make the monitors for laptops do they not? Freesync is not that different from what was already being done with laptop screens right?
Just because you haven't heard of it doesn't mean it's not true.I've not heard of it being used for multi-socket communication (inter-CCX within a CPU yes) but it's definitely not something communicating with the GPU.