NVIDIA today unveiled the GeForce RTX 30 series of graphics cards which utilize third-generation Tensor cores and are, themselves, the company’s second-generation attempt at consumer-level ray tracing rendering. Not that the RTX 20 series wasn’t pretty accomplished already, but the size of this forward step that we are now witnessing is truly astonishing.
To the point that it would probably be naive to expect to see any of these newly introduced GPUs in stock until at least spring, as is the unfortunate tradition. But either way, the new Turing architecture used by the GeForce RTX 30 line offers single-precision floating-point format throughput that’s twice as fast. That’s a lot of extra frames per second you can expect from Cyberpunk 2077.
Even if you disregard the flagship GeForce RTX 3090 which sweeps the floor with pretty much every GPU that ever existed. Even the unassumingly mid-range RTX 3070 promises to at the very least match – and in some cases, exceed – the performance of an RTX 2080 Ti. And that’s a $499 GPU we’re talking about, compared to a $999 one. The GeForce RTX 3080 and 3090 obviously bring things up a notch, but it remains to be seen how soon after their official release dates they actually become available for purchase to us, regular mortals.
Speaking of, the RTX 3080 should be out in the U.S. on September 17th, the 3090 is coming exactly a week later, and the 3070 is currently planned for an October debut.