Update: Gigabyte themselves have come out of the gates dismissing this as a typo on their part, which is disappointing, if not unexpected, considering that there is no real reason for NVIDIA to launch a new SKU to a market virtually absent of competition. Increased tiers of graphics card just give more options for the consumer, and why give an option that might drive customers away from more expensive graphics card options?

NVIDIA designed the $500 GeForce RTX 2070 based on its third largest silicon based on "Turing," the TU106. Reviews posted late Tuesday summarize the RTX 2070 to offer roughly the the same performance level as the GTX 1080 from the previous generation, at the same price. Generation-to-generation, the RTX 2070 offers roughly 30% more performance than the GTX 1070, but at 30% higher price, in stark contrast to the GTX 1070 offering 65% more performance than the GTX 970, at just 25% more price. NVIDIA's RTX varnish is still nowhere in sight. That said, NVIDIA is not on solid-ground with the RTX 2070, and there's a vast price gap between the RTX 2070 and the $800 RTX 2080. GIGABYTE all but confirmed the existence of an SKU in between.Called the GeForce RTX 2070 Ti, this SKU will likely be carved out of the TU104 silicon, since NVIDIA has already maxed out the TU106 with the RTX 2070. The TU104 features 48 streaming multiprocessors in comparison to the 36 featured by TU106. NVIDIA has the opportunity to carve out the RTX 2070 Ti by cutting down SM count of the TU104 to somewhere in the middle of those of the RTX 2070 and the RTX 2080, while leaving the memory subsystem untouched. With these, it could come up with an SKU that's sufficiently faster than the GTX 1070 Ti and GTX 1080 from the previous generation, and rake in sales around the $600-650 mark, or exactly half that of RTX 2080 Ti.