@nomadic8280 said:
@ronvalencia said:
@nomadic8280 said:
@zaryia said:
Probably 20-30 fps again. The casuals care more about teh resolution even though they can't notice it as much as FPS due to TV distance. It's mostly marketing and what not.
Fortunately we have PC as an option.
Well hold on now. By next gen, 4K Tv's will just barely be taking over as commonplace in most homes (if that). There won't be 8K for a looooong time, the time will be right for the hardware to deliver native 4K @ 60 FPS. Can't see another big feature for the next gen.
For 4K 60 fps game consoles, it needs to be atleast twice of X1X's hardware e.g. 12 TFLOPS and 652 GB/s memory bandwidth (assuming there's no improvements from Polaris delta color compression). CPUs needs to scale by 2X.
GTX 1080 Ti has 12.9 TFLOPS at 1800Mhz and it's close to 4K 60 fps for most X1X ports.
GTX 1080 level that replaces RX-580/RX-580X in 7nm process tech is not enough. MS may have to build another Xbox One X like beast for 7 nm generation that exceeds RX-680.
Xbox One X's PCB has 384 bit trace lines which is a high end PC PCB design.
X1X's GPU solution exceeds RX-580 solution.
FreeSync reduces the need for locked 60 hz.
It's not looking good for a next gen $400-500 box that will do anything that much better than the Pro or X then, by 2020 anyway. To invest in an entire new product, they need to show a clear night and day difference to get consumers out the door to buy it. "You know how some games were 4K low settings, 30fps? Now they're going to be 4K ultra settings, 30fps"...that just won't cut it.
Far Cry 5 and Gears of War 4 are not low settings on X1X.
Next progress is with DirectX Ray-tracing (DXR) over PBR (Physical Based Rendering).
https://blogs.msdn.microsoft.com/directx/2018/03/19/announcing-microsoft-directx-raytracing/
DXR is compute like workload which plays into AMD's strength on mainstream $250 cards against similar NVIDIA's $250 mainstream cards.
From MS's blog
You may have noticed that DXR does not introduce a new GPU engine to go alongside DX12’s existing Graphics and Compute engines. This is intentional – DXR workloads can be run on either of DX12’s existing engines. The primary reason for this is that, fundamentally, DXR is a compute-like workload. It does not require complex state such as output merger blend modes or input assembler vertex layouts. A secondary reason, however, is that representing DXR as a compute-like workload is aligned to what we see as the future of graphics, namely that hardware will be increasingly general-purpose, and eventually most fixed-function units will be replaced by HLSL code. The design of the raytracing pipeline state exemplifies this shift through its name and design in the API. With DX12, the traditional approach would have been to create a new CreateRaytracingPipelineState method. Instead, we decided to go with a much more generic and flexible CreateStateObject method. It is designed to be adaptable so that in addition to Raytracing, it can eventually be used to create Graphics and Compute pipeline states, as well as any future pipeline designs.
"Output merge blend modes" and "fixed function units" deals with ROPS hardware and it's less on DXR.
With very little software, AMD jump the gun on TFLOPS compute bias GPU design. Prepare for another Finewine advantage for AMD.
Larger NVIDIA GPUs such as GTX 1080 and 1080 Ti has high TFLOPS comparable to Vega 56/64s.
There's a high probability, DXR will smash GTX 1060s into the ground i.e. Kepler style aging.
NVIDIA is promoting Titan V due to it's 15 TFLOPS FP35 compute shader for DXR. NVidia's tensor units are useless for Direct3D APIs.
Log in to comment