Rumors about NVIDIA Ampere: more power for ray tracing, higher frequency and more memory

According to rumors, the next generation of NVIDIA GPUs will be called Ampere, and today WCCFTech shared a large portion of unofficial information regarding these chips and video cards based on them. It is reported that the following information was reported by NVIDIA to its partners, so it should be very reliable.

Rumors about NVIDIA Ampere: more power for ray tracing, higher frequency and more memory

The first thing NVIDIA plans to focus on with Ampere GPUs is ray tracing. The company promises that GeForce RTX 30-series graphics cards will provide a significant performance boost in ray tracing over current GeForce RTX 20-series solutions. The RT cores responsible for tracing in the Ampere architecture will be faster and more power efficient, and there will simply be more of them compared to Turing.

Also in the Ampere architecture, NVIDIA wants to improve rasterization performance. NVIDIA has long been paying increased attention to this area, due to which its GPUs often outperform AMD solutions when processing complex geometry. Initially, the emphasis on rasterization performance was placed in professional Quadro accelerators, but now GeForce consumer cards can get significant improvements in this direction.

Rumors about NVIDIA Ampere: more power for ray tracing, higher frequency and more memory

It is noted that the complexity of game worlds is growing and increased performance during rasterization will enable the next generation of NVIDIA GPUs to work with them with greater efficiency. But in general, both rasterization and ray tracing will be of the greatest importance in games after the release of next-generation consoles, so NVIDIA is probably moving in the right direction.

The source also provides information about the characteristics of future video cards, though only in general terms, without any specific figures. First, it is reported that Ampere GPUs will have a larger frame buffer compared to Turing. That is, the amount of video memory will increase.

Secondly, the transition to a 7-nm process technology (7 nm EUV) will increase the frequency of the chips by about 100-200 MHz. Also, due to the transition to a “thinner” process technology, Ampere GPUs will operate at a lower voltage, most likely below 1 V. This can potentially reduce the overclocking potential of the chips. But at the same time, this will increase the energy efficiency of new video cards.

Rumors about NVIDIA Ampere: more power for ray tracing, higher frequency and more memory

And finally, it is reported that the cost of NVIDIA graphics cards based on Ampere GPUs will be about the same as graphics cards based on Turing chips. Perhaps older solutions, conditional GeForce RTX 3080 and RTX 3080 Ti, may cost less than their predecessors. True, it is important to note here that it is still too early to talk about the cost, because many factors can affect it. Ampere generation video cards should be released next year.



Source: 3dnews.ru

Add a comment