NVIDIA made a splash as the first major press conference of CES 2019. In addition to a new and mercifully cheap addition to its top-of-the-line RTX graphics card family, the company announced RTX-series chips for gaming laptops. Both will be here before the end of the month.
The RTX 2060 card is based on the same Turing GPU as the RTX 2070 and 2080 cards that debuted last year, but with a much more attainable price. At $350 for NVIDIA’s self-branded card (a bit higher or lower for tweaked designs from partner manufacturers like Gigabyte and EVGA), it’s less than half the price of the most expensive option, and this model will be making its way to the interior of many a gamer’s desktop PC. It should also be a popular option for pre-built models. Even more frugal cards, undoubtedly including 2050 and 2050ti variants, should come later this year.
NVIDIA says the RTX 2060 is more powerful than the GTX 1070ti (first released in November 2017) for most modern games. It comes with 6GB of GDDR6 RAM and can handle five gigarays of computation per second. It will be available starting on January 15th.
In addition to the desktop side, NVIDIA also pulled the wraps off of the first laptops to feature RTX-branded chips. Laptops from major manufacturers with RTX 2060, 2070, and 2080 GPUs will begin hitting the market on January 29th. These include both standard models and those with NVIDIA’s Max-Q technology, allowing for thinner and more power-efficient designs with some extra special sauce in the thermal management. 40 RTX-packing laptops are in the works, among them 18 Max-Q laptops, including an update to the popular MSI G65 Stealth Thin, which Review Geek recommended highly. Expect RTX cards to be limited to the most expensive laptops, at least for the first few months.
The biggest draw for the RTX line of desktop and mobile cards is ray tracing capability, an option for super-realistic lighting and reflections. Early RTX buyers were discouraged at the performance hit in brand new games when RTX lighting was enabled, often driving games down below the magical 60 frames-per-second mark even on massive powerhouse desktops. NVIDIA’s been working on the problem and says that improved drivers featuring DLSS (Deep Learning Super-Sampling) on more compatible games will boost that performance.
According to the presentation, DLSS should close the performance gap to within one or two frames per second of the game running with RTX disabled. DLSS combines upscaled sampling with NVIDIA’s proprietary AI technology to smooth out the rough patches of the GPU-intensive ray tracing process. We should see more games support it at launch as RTX ray tracing becomes standard on new high-end PC games.