NVIDIA today launched its GeForce RTX 20-series graphics card family based on its ambitious new “Turing” architecture. Launched 18 months from “Pascal”, Turing comes at a time when advancements in silicon fabrication node technology are unable to keep pace with roadmaps of major chipmakers who traditionally brought out a new architecture based on a new process every 18–24 months.
In an ideal world, we should have gone sub-10 nm already, which NVIDIA would have leveraged to bring the “Volta” architecture to the consumer-space for another serving of “more of everything.” The “Turing” architecture packs a collection of innovations that were needed to build a new GPU on existing silicon fab processes.
At the heart of NVIDIA’s effort is the RTX Technology, which brings what looks like real-time ray tracing to 3D games. Not everything on your screen is ray traced, but some of the objects are; and so, a hybrid of ray tracing and classic rasterization makes up what you see.
To ray trace even those few things on your screen, an enormous amount of compute power is needed, and so NVIDIA created specialized hardware for the task in the form of RT cores, which sit besides the all-purpose CUDA cores. The Tensor cores, which made their debut with “Volta”, also feature here, lending a hand with deep learning and AI tasks, including a few turnkey features game developers can integrate. The new architecture also keeps up with generational gains in memory bandwidth with the new GDDR6 memory standard. The display I/O is revamped with support for the latest DisplayPort and HDMI standards, and a revolutionary new connector called VirtualLink.
In its long list of firsts, “Turing” also sees NVIDIA debut the architecture not with two SKUs based on the second-biggest chip (e.g.: GTX 1080 and GTX 1070), but the flagship SKU based on the “big chip”, along with the top SKU based on the second-biggest chip, with the introduction of the new GeForce RTX 2080 Ti and GeForce RTX 2080. The GeForce RTX 2070 is also on the horizon, but isn’t launching today. NVIDIA is saving its launch for next month.
The GeForce RTX 20-series is launching at unusually high prices, with generational price increments ranging between 15%–70%. NVIDIA’s justification is that these cards are “more than GeForce GTX”, and has made a few tweaks to its product stack. The RTX 2070, which starts at $500, is the cheapest SKU for now, followed by the RTX 2080 at $700 and the flagship RTX 2080 Ti at $1000, at least.
These prices don’t apply to “reference design” cards, which don’t quite exist. Cards that are completely designed by NVIDIA are referred to as “Founders Edition”, which not only sell with a premium product design, but higher-than-reference clock speeds to justify 10%–15% premiums.
In this review, we have with us the MSI GeForce RTX 2080 Ti Duke. The company originally designed the Duke series as step down to its Gaming X series in the product stack. These cards typically combine a close-to-reference PCB with a custom-design triple-fan cooling solution and a minor factory overclock to sweeten the deal. With the RTX 20-series, however, MSI has promoted the Duke brand to have a slightly more premium aesthetic, which includes premium TorX 2.0 fans, a more pronounced RGB LED ornament, and a backplate, not to mention an 8 percent factory overclock for 1350 MHz GPU clocks and 1665 MHz GPU Boost.
Our exhaustive coverage of the NVIDIA GeForce RTX 20-series “Turing” debut also includes the following reviews:
NVIDIA GeForce RTX 2080 Ti Founders Edition 11 GB | NVIDIA GeForce RTX 2080 Founders Edition 8 GB | ASUS GeForce RTX 2080 Ti STRIX OC 11 GB | ASUS GeForce RTX 2080 STRIX OC 8 GB | Palit GeForce RTX 2080 Gaming Pro OC 8 GB | MSI GeForce RTX 2080 Gaming X Trio 8 GB | MSI GeForce RTX 2080 Ti Gaming X Trio 11 GB | NVIDIA RTX and Turing Architecture Deep-dive
Price | Shader Units |
ROPs | Core Clock |
Boost Clock |
Memory Clock |
GPU | Transistors | Memory | |
---|---|---|---|---|---|---|---|---|---|
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 |
MSI RTX 2080 Ti Duke | $1199 | 4352 | 64 | 1350 MHz | 1665 MHz | 1750 MHz | TU102 | 18600M | 11 GB, GDDR6, 352-bit |