NVIDIA launched the GeForce RTX 20-series with the introduction of the GeForce RTX 2080 and RTX 2080 Ti. It comes at a time when the silicon fabrication technology isn't advancing at the rate it used to four years ago, wrecking the architecture roadmaps of several semiconductor giants, including Intel, NVIDIA, AMD, and Qualcomm, which is forcing them to design innovative new architectures on existing foundry nodes. Brute transistor-count increases, as would have been the case with "Volta," are no longer a viable option, and NVIDIA needed a killer feature to sell new GPUs. That killer feature is the RTX Technology. This feature is so big for NVIDIA that it has changed the nomenclature of its client-segment graphics cards with the introduction of the GeForce RTX 20-series.
NVIDIA RTX is a near-turnkey real-time ray-tracing model for game developers that lets them fuse real-time ray-traced objects into 3D scenes that have been rasterized. Ray-tracing the whole scene in existence isn't quite possible yet, but the results with using RTX are still better-looking than anything rasterizing can achieve. To even get those few bits of ray tracing done right, an enormous amount of compute power is required. NVIDIA has hence deployed purpose-built hardware components on its GPUs that sit alongside all-purpose CUDA cores, called RT cores.
NVIDIA invested heavily to stay at the bleeding edge of the hardware that drives pioneering AI research and has, over the years, developed Tensor cores, specialized components that are tasked with matrix multiplication, which speeds up deep-learning neural-net building and training via Tensor ops. Although it's a client-segment GPU for gaming, NVIDIA feels GPU-accelerated AI could play an increasingly big role in the company's turnkey GameWorks effects and a new image quality enhancement called Deep-Learning Super-Sampling (DLSS). The chips are hence endowed with Tensor cores, just like the TITAN Volta. All that it lacks compared to the $3,000 graphics card from last year is FP64 CUDA cores.
NVIDIA GeForce RTX 20-series graphics cards debut at unusually high prices compared to their predecessors, perhaps because NVIDIA doesn't count the GTX 10-series as a predecessor to begin with. These chips pack not just CUDA cores, but also RT cores and Tensor cores, adding to the transistor count which, along with generational increases in performance, contributes to scorching 15%–70% increases in launch prices over the GTX 10-series. The GeForce RTX 2080 is the second-fastest graphics card from the series and is priced at $700 for the base model.
EVGA's RTX 2080 Ti XC Ultra is based on the NVIDIA reference design PCB, but comes with a much more capable triple-slot, dual-fan thermal solution. In terms of clocks, EVGA has bumped Boost to 1650 MHz, which is 25 MHz higher than the Founders Edition. Pricing is a bit higher than the Founders Edition: $1,250.
|GTX 1070||$390||1920||64||1506 MHz||1683 MHz||2002 MHz||GP104||7200M||8 GB, GDDR5, 256-bit|
|RX Vega 56||$400||3584||64||1156 MHz||1471 MHz||800 MHz||Vega 10||12500M||8 GB, HBM2, 2048-bit|
|GTX 1070 Ti||$400||2432||64||1607 MHz||1683 MHz||2000 MHz||GP104||7200M||8 GB, GDDR5, 256-bit|
|GTX 1080||$470||2560||64||1607 MHz||1733 MHz||1251 MHz||GP104||7200M||8 GB, GDDR5X, 256-bit|
|RX Vega 64||$570||4096||64||1247 MHz||1546 MHz||953 MHz||Vega 10||12500M||8 GB, HBM2, 2048-bit|
|GTX 1080 Ti||$675||3584||88||1481 MHz||1582 MHz||1376 MHz||GP102||12000M||11 GB, GDDR5X, 352-bit|
|RTX 2070||$499||2304||64||1410 MHz||1620 MHz||1750 MHz||TU106||10800M||8 GB, GDDR6, 256-bit|
|RTX 2070 FE||$599||2304||64||1410 MHz||1710 MHz||1750 MHz||TU106||10800M||8 GB, GDDR6, 256-bit|
|RTX 2080||$699||2944||64||1515 MHz||1710 MHz||1750 MHz||TU104||13600M||8 GB, GDDR6, 256-bit|
|RTX 2080 FE||$799||2944||64||1515 MHz||1800 MHz||1750 MHz||TU104||13600M||8 GB, GDDR6, 256-bit|
|RTX 2080 Ti||$999||4352||64||1350 MHz||1545 MHz||1750 MHz||TU102||18600M||11 GB, GDDR6, 352-bit|
|RTX 2080 Ti FE||$1199||4352||64||1350 MHz||1635 MHz||1750 MHz||TU102||18600M||11 GB, GDDR6, 352-bit|
|EVGA RTX 2080 Ti XC Ultra||$1249||4352||64||1350 MHz||1650 MHz||1750 MHz||TU102||18600M||11 GB, GDDR6, 352-bit|