قالب وردپرس درنا توس
Home / Technology / NVIDIA's RTX Card is a gamble on the future of games

NVIDIA's RTX Card is a gamble on the future of games



Due to rough power, the RTX 2070 is approximately equivalent to GTX 1080; The RTX 2080 goes toe to the toe with the GTX 1080 Ti; The RTX 2080 Ti is in its own league. 2070 and 2080 have 8GB of GDDR6 RAM; 2080 Ti has 11 GB. All three are based on the company's new Turing architecture, which means that they have cores dedicated to AI (Tensor) and Ray-Tracing (RT).

Expect a fourth card, probably the RTX 2060, to significantly lower the entry price in the coming months, followed by a number of budgeted cuts options (1

0 Series dropped to sub- $ 100 GTX 1030). There is also space on the top for expansion: The RTX 2080 Ti Founders Edition can handle 14.2 trillion floating point operations per second (TFLOPS), while the Turing TU102 chip these new cards are based on pushes it up to 16.3 TFLOPS. It is achieved through a mix of higher clock speed and more CUDA cores (2080 Ti has 4,352, the fully configured TU102 has 4,608.)

RTX also comes with many improvements under the cover. It is a faster caching system with a shared memory architecture, a new graphics pipeline and simultaneous processing of floating and integer calculations. If that does not matter to you, do not worry too much: Take-away from that word is not only, the RTX series has more raw power, but it uses that power more efficiently.

And that's the key here. Ray tracing stole the headlines, and I'm thrilled to see how developers use it, but it's efficiency that really excites me about RTX.

The ultimate goal of a gaming system, be it a $ 2000 play PC or a $ 300 Nintendo Swap, is to calculate a color value for each pixel on a screen. Even a simplified guide to how a modern graphics pipeline does this will run the length of a novella, but here's a three-sentence summary: CPUs are not designed to render modern graphics. Instead, a CPU sends a plan for what it wants to draw to a GPU, which has hundreds or thousands of cores that can work independently of small pieces of an image. The GPU performs on the CPU plan, running shaders – very small programs – to define the color of each pixel.

The challenge for both video card manufacturers and game developers is then the scale. The $ 300 switch, in portable mode, usually calculates 27 million pixel values ​​a second, as it works well with a three-year-old mobile NVIDIA chip. If you target 4K at 60FPS (as many players buy RTX cards), the system must move close to half a billion pixels per second. It puts a lot of stress on a system, especially when considering that the PC not only picks these colors out of thin air, and instead simulates a complex 3D environment in real time as part of the calculations.

There are already many techniques used to reduce this load. It reproduces all or part of a lower resolution scene and extends the results. This is super obvious when you have a 720p drive on a 1080p screen, but less so when it is said, it fails a fog on quartet resolution. And that's what NVIDIA's optimization is about: cut down the quality of places you will not notice.

NVIDIA's new graphics pipeline can use more new shading techniques to cut corners. In many ways, this builds on less flexible energy-saving measures that are used for VR, such as MRS (multi-resolution shading) and LMS (lens shading). In the image above you will see a GPU that breaks down a scene in a real-time grid. The uncolored squares are high-definition and shaded in the ratio 1: 1, just like a regular game scene. The colored ones do not need the same attention. The red squares, for example, are only shaded in 4×4 pixel blocks, while more detailed but non-essential blue squares are shaded in 2×2 blocks. Given the low level of detail for the areas of the image, the change is essentially unnoticeable.


Source link