Baseless Next Generation Rumors with no Technical Merits [post E3 2019, pre GDC 2020] [XBSX, PS5]

I mean if it’s that easy, I have a lot of gadgets at home.

 

I don't think the chances are very high for the screen, but here's a 2 inch 320x240 screen.... It's $2 in 1000 qty.

https://www.alibaba.com/product-detail/240-X-320-2-0-Inch_62377221891.html

And that 1" oled on the ds4 button adapter is below $1 in large qty.

 

The saltyness with this is awsome,

For thoes who don't know: a company made a very black paint and Anish brought the exclusive rights to it. No other artist was allowed to use "his" ultra black black. This product is a direct response to that with added snark.

*Note: By adding this product to your cart you confirm that you are not Anish Kapoor, you are in no way affiliated to Anish Kapoor, you are not purchasing this item on behalf of Anish Kapoor or an associate of Anish Kapoor. To the best of your knowledge, information and belief this material will not make it's way into the hands of Anish Kapoor.

 

I don't exclude the possibility of early communication with AMD from the start, I merely conclude that given the rapid deployment of NVIDIA's RT initiative across the board, it's efforts most probably predated DXR. Meaning they were already in the process of doing hardware accelerated RT in the form of a CUDA/Optix extension, and later as a DX12/Vulkan proprietary extension. This is the only explanation for their seemingly immediate hardware support for DXR.

DXR was first announced in March 2018, it was immediately supported on Volta (which was released in May 2017), with announced support for later architectures, which turned out to be Turing (just an upgraded Volta + RT cores), Turing got released a few months later in August 2018. Things don't happen this quickly unless NVIDIA has laid the groundwork for hardware RT before DXR, especially as AMD is completely absent from that picture, didn't even come up with an official hardware RT solution to this day, meaning they probably started working on RT during the time of DXR formulation, not before it.

Yeah, they practically got 4 DX12 features early, however these features were also ahead of their DX12 integration, meaning NVIDIA was probably independently developing them already.

They had AI hardware acceleration through Tensor cores on CUDA, which got integrated later into DirectML (which is still in it's beta phase).
They had Mesh Shaders on Turing on a separate DX12/Vulkan proprietary extension, which is still currently being integrated into DX12.
They even had Variable Rate Shading (VRS) on Turing since day one, through a separate Vulkan proprietary extension, DX12 began integrating the feature some 8 months after Turing announcement.

The last one and the big one is Ray Tracing, can you spot the pattern? I am postulating NVIDIA had RT acceleration even before DXR, at least in the same way they had AI acceleration before DirectML.

If Sony is really developing a custom RT solution, then Sony's RT effort most probably have predated DXR, same as NVIDIA.

 

Anybody consider that RT was already planned for next gen consoles and Nvidia accelerated its RT plans to get a jump on the market? 12nm is not the ideal node for offering RT while trying to offer more compute/rasterization performance than previous gen.

RTX seemed extremely rushed especially considering that if RT is hard to implement as a good design and AMD had no RT on the horizon planned, then Nvidia could have easily waited until 7nm. 20 series card at 7nm make a lot more sense than at 12nm for almost all other variables other than time.

 

This sounds like a whole lot of invitations to problems coming to their doorstep.
- Of course he won, the A.I. was practically moving for him!
- Of course I lost, the stupid A.I. kept moving me to the wrong place!

 

Both are valid POVs.
I just don’t agree with the idea that the regression tests help provide evidence that MS is using AMD and Sony is not.

the tests on RT in my opinion show nothing; Except that a test could be run. The results are inconclusive.

 

Do you have a full unedited picture of this one?

 

Why not just use your existing Google spy-gadget?

 

If HW RT is only for ray intersection, how are they implementing ray marchers in real time?

 

Should be the special sauce RT chip...

 

This is interesting. I have speculated that Ariel and Gonzalo (characters from Tempest) are SOC and platform for dev kit (V style), while Oberon/Flute (characters from A Midsummer Night's Dream) are same thing for retail console.

 

It's hard to rush a complete integration of something 4 or 5 years in advance, NVIDIA's RT efforts must have started in 2015 or 2014, (when they were planning for Volta/Turing), they were also heading toward RT since their days of accelerating RT rendering through CUDA on Fermi (2010), then Real Time RT demos on Kepler (2013), then doing some partial RT effects like HFTS, VXAO on Maxwell and Pascal.

 

There will be different gadgets across the house to talk to your voice assistant. The console will just be one of them. I know two persons who have bough 2 google home speakers, one for the bedroom and the other one for the living room (+ their phone).

In the future we'll have hundreds of mics in our houses so the NSA can chose to tap which ever they want.

 

At the moment, I'd rather spy him instead...

 

Look, look...

13F9:OBR:A0



So its now confirmed that Flute benchmark used Oberon chip.

Now, check out what Taiwanese leaker AquariusZi said back in October :

Now check socket in that Flute benchmark...

 

This one ?

 

*squinting*

13.9 TF OBERAON?

 

VLIW TFs!