Nvidia Turing architecture deep dive
Nvidia announced three new GeForce RTX graphics cards at Gamescom 2018, the GeForce RTX 2080, GeForce RTX 2080 Ti, and GeForce RTX 2070. We also have full reviews of the GeForce RTX 2080 Ti Founders Edition and GeForce RTX 2080 Founders Edition. We've covered all the details of what you need to know about each GPU in those articles, but for those that want to get technical, this will be an in-depth look at the Nvidia Turing architecture.
There's a ton of new technology in the Turing architecture, including ray tracing hardware, and Nvidia rightly calls it the biggest generational improvement in its GPUs in more than ten years, perhaps ever. If the specifications for Turing didn't seem like anything special, just the same old stuff with some fancy ray tracing and deep learning marketing slapped on, there's far more going on than paper specs. If you want the lowdown on all of Turing's deepest and darkest secrets, you've come to the right place.
Nvidia's last GeForce architecture was Pascal, which powered everything from the top-tier best graphics cards like the GeForce GTX 1080 and GeForce GTX 1080 Ti to the entry level GeForce GTX 1050 and GeForce GT 1030. Last year, Nvidia released its new Volta architecture, adding Tensor cores and HBM2 memory to the picture, but the only 'pro-sumer' card to use Volta is the Titan V, a $2,999 monster. The Volta GV100 will apparently remain in the supercomputing and deep learning focused fields, because the new Turing architecture appears to beat it in nearly every meaningful way. If you just splurged on a Titan V, that's bad news, but for gamers holding out for new graphics cards, your patience (and common sense) has paid off.
Full Turing architecture specs
There was a ton of speculation and, yes, blatantly wrong guesses as to what the Turing architecture would contain. Prior to the initial reveal at SIGGRAPH 2018, every supposed leak was fake. Making educated guesses about future architectures is a time honored tradition on the Internet, but such guesses are inevitably bound to be wrong. We've covered many details of Nvidia's Turing architecture previously, but today we can finally take the wraps off everything and get into the minute details.
I'm including both the 'full' Turing chips as well as the GeForce RTX Founders Edition variants in the following table, plus the previous generation Pascal equivalents. The 20-series Founders Editions have a 90MHz higher boost clock, putting them in the same range that factory overclocked models are likely to land. Here are the complete Turing specs, including die size and transistor counts for the smaller Turing cores:
Compared to the previous generation Pascal parts, the Turing architecture has similar clockspeeds but increases CUDA core counts by 15-20 percent across the line. That's only the beginning, as Nvidia also adds Tensor cores and RT cores to the picture, and the individual SMs (streaming multiprocessors) have seen significant changes. More on that in a moment.
Another big change is that Nvidia is launching three fully separate GPU dies for the high-end and enthusiast segment at the same time. Previously, the 1070/1080 and 970/980 were built from the same die, with the lesser part using a partially disabled version. The 2080 and 2080 Ti still use harvested dies, but the 2070 gets a separate and complete TU106 GPU. That also leaves room for future in-between GPUs like a 2070 Ti and Titan RTX, naturally.
Pure memory bandwidth also sees a healthy improvement, thanks to GDDR6. The 2080 Ti sees the smallest improvement of 27 percent, since the 1080 Ti already used 11 GT/s GDDR5X. The reference 1080 uses 10 GT/s GDDR5X (some later custom models use 11 GT/s VRAM) so the RTX 2080 has 40 percent more bandwidth. The potential winner is the RTX 2070, which gets a substantial 75 percent boost in bandwidth.
But bandwidth and theoretical performance are only part of the equation. Let's dive into deeper waters and talk about the low level details of the GPU, memory interface, and other aspects of the Turing architecture.
Turing architecture: the GPU core
Above is the full block diagram for the Turing TU102/TU104/TU106 architectures. The TU102 consists of six GPCs (Graphics Processing Clusters), each of which contains six TPCs (Texture Processing Clusters), a PolyMorph engine, and a dedicated rasterization engine. Each TPC is in turn linked to two SMs (Streaming Multiprocessors). Is that enough acronyms to get us started?
Along with the GPCs, at a high level the TU102 includes 12 32-bit GDDR6 memory controllers (384-bit total), which can be independently disabled. The memory controllers also contain the ROPs (Render Outputs), so the RTX 2080 Ti with 11 memory controllers also ends up with 88 ROPs.
The PCIe 3.0 host interface and other elements also reside outside the GPCs. Note that the above diagrams should also not be taken as a literal representation of how the chips are laid out on silicon but are merely a high level overview.
It's interesting that the GPCs are not uniform across all the Turing architecture GPUs. The TU104 also has six GPCs, but each GPC in the TU104 has eight SMs where the TU102 and TU106 GPCs have 12 SMs. The TU104 and TU106 both have eight 32-bit memory controllers (256-bit total), however, along with the various other functional units.
At the heart of every GPU is the fundamental building block. Nvidia calls this the Streaming Multiprocessor (SM) while AMD calls it a Compute Unit (CU), but while the specific implementations vary, each GPU typically has many clusters of SMs.
The Turing architecture SMs contain schedulers, graphics cores, L1/L2 cache, texturing units, and more. Nvidia has dramatically altered the Turing architecture SM compared to the previous Pascal and Maxwell architectures, so there's a lot to cover. Let's start at the top.
First, the number of CUDA cores per SM is now 64 instead of Pascal's 128. Nvidia has bounced around over the past decade with anywhere from 32 to 192 CUDA cores per SM, but Nvidia says with the other architectural changes 64 cores is now more efficient.
The Turing architecture also adds native 'rapid packed math' FP16 support to the CUDA cores, which was previously seen in GP100 and GV100. Performance for FP16 workloads is double that of the FP32 cores, though games predominantly use FP32. Not shown in the above SM block diagram are the FP64 CUDA cores, which are separate from the FP32 cores. There are two FP64 cores per SM, for compatibility purposes, so FP64 performance is 1/32 the FP32 performance. (Volta GV100 and Pascal GP100 both have half-speed FP64 support, which is useful in many supercomputing workloads.)
New for the Turing architecture is a dedicated integer pipeline that can run concurrently with the floating point cores. While graphics applications predominantly use FP calculations, Nvidia has profiled a large number of games and other applications and said that typically there are 35 (or 36, depending on which Nvidia document you consult) integer instructions for every 100 FP instructions.
On previous architectures, the FP cores would have to stop their work while the GPU handled INT instructions, but now the scheduler can dispatch both to independent paths. This provides a theoretical immediate performance improvement of 35 percent per core.
This makes the GPU cores in Turing more like modern CPU architectures, and the scheduler can dispatch two instructions per clock cycle. These instructions can also be for the RT cores and Tensor cores.
Ray tracing with the new RT cores - full size image.
Swipe for how ray tracing works on Pascal.
Ray tracing using shaders on Pascal - full size image.
Swipe for how ray tracing works on Turing.
Turing architecture: ray tracing with RT cores
Most of the above items were present in Nvidia's previous GPU architectures. The Turing architecture brings two new capabilities, starting with RT cores for ray tracing. I've covered ray tracing in more detail elsewhere, so this is the condensed version focused on the architecture side of things.
Each Turing SM now adds a single RT core. Nvidia doesn't provide an exact performance number, since the actual ray tracing BVH algorithm isn't deterministic, meaning it doesn't always execute in the same amount of time. Nvidia says the RT cores do ">10 Giga Rays per second" (GR/s) in the GeForce RTX 2080 Ti, and that it takes about 10 TFLOPS of computations for each GR/s. Working back from that to the clockspeeds and RT core counts, I've estimated the 'exact' GR/s performance for the various Turing GPUs in the above specs table.
It's important to state that these RT TFLOPS are not general purpose TFLOPS, but instead these are specific operations designed to accelerate ray-tracing calculations. The RT cores compute ray triangle intersections (where a ray hits a polygon), as well as BVH traversal. That second bit requires a lengthier explanation.
BVH stands for "bounding volume hierarchy" and is a method for optimizing intersection calculations. Instead of checking rays against polygons, objects are encapsulated by larger, simple volumes. If a ray doesn't intersect the large volume, no additional effort needs to be spent checking the object. Conversely, if a ray does intersect the bounding volume, then the next level of the hierarchy gets checked, with each level becoming more detailed.
All of the BVH calculations can be done using shader cores, but Nvidia says it takes thousands of shader calculations per ray, during which time the CUDA cores can't be working on other stuff. The RT cores offload all of that and run independently with the CUDA cores, so on the Turing architecture enabling ray tracing won't completely tank performance.
The final major architectural feature in Turing is the inclusion of Tensor cores. Normally used for machine learning, you might wonder why these are even useful for gaming. I'll have a separate piece digging deeper into the machine learning aspects of Turing, but in short there's a lot of future potential.
Nvidia has worked with Microsoft to create the DirectML and Windows ML (DirectX Machine Learning) APIs, so this is something with broad industry support. Future games could use machine learning to enhance AI in games, offer improved voice interfaces, and enhance image quality. Those are all longer term goals, however—especially when for the next five or more years a large installed base of gamers won't have Tensor cores available. In the more immediate future, these cores can be used in more practical ways.
Nvidia showed some examples of improved image upscaling quality, where machine learning that has been trained on millions of images can generate a better result with less blockiness and other artifacts. Imagine rendering a game at 1080p with a high framerate, but then using the Tensor cores to upscale that to a pseudo-4k without the massive hit to performance we currently incur. It wouldn't necessarily be perfect, but suddenly the thought of 4k displays running at 144Hz with 'native' 4k content isn't so far-fetched.
The Tensor cores are also required to handle an AI-trained denoising algorithm for ray tracing. While the Tensor cores are running, the rest of the GPU basically ends up being idle, so unlike the RT cores and INT/FP pipelines the Tensor cores don't really work concurrently. However, Nvidia suggests that DLSS and denoising could run with only 20 percent of the total frame time.
At a hardware level, the Turing architecture Tensor cores also add a few new tricks compared to those in Volta. Specifically, native support for INT8 and INT4 workloads allows for potentially double or quadruple the computational performance relative to FP16/FP32 hybrid models. These may not be immediately applicable to graphics workloads, which are more susceptible to quantization, but research into alternative machine learning algorithms is ongoing.
RTX-OPS, a new performance metric
With all of the changes in the Turing architecture, comparing performance between various GPU generations just became a lot more difficult. For existing games and workloads that don't leverage the new features, the old FP32 TFLOPS figure might still be somewhat okay, but the superscalar design (ie, the ability to dispatch multiple concurrent instructions) muddies even those figures. To help compare things, Nvidia devised a new performance metric, RTX-OPS.
Obviously this will favor the RTX GPUs, but the general idea isn't that bad. In a modern game that fully implements ray tracing effects with DLSS and denoising, the above slide shows how the average workload gets distributed. 80 percent of the time is spent on FP32 shading (what games currently do), with 35 percent of that time also having concurrent INT32 shading work. Ray tracing calculations meanwhile use another 40 percent of the time, and the final output gets post-processed by the Tensor cores taking a final 20 percent.
With that formula in hand, the GeForce RTX 2080 Ti FE ends up with 78 RTX-OPS, as shown. How would a GTX 1080 Ti compare? It lacks the Tensor cores and RT cores and can't do simultaneous INT32 + FP32 calculations, so everything uses the base TFLOPS figure and just uses a different slice of the overall pie. In other words, GTX 1080 Ti has RTX-OPS equal to its TFLOPS figure of 11.3.
Does that mean the RTX 2080 Ti is nearly seven times faster than the 1080 Ti? Not really, but in future workloads that use ray tracing and machine learning code, previous architectures simply won't be able to compete.
Turing architecture: GDDR6 and improved L1/L2 cache
Improving overall GPU performance is great, but faster GPUs also need more memory bandwidth. To keep the GPUs fed with data, Nvidia has moved to GDDR6 memory and reworked the cache and memory subsystem in the Turing architecture. L1 cache bandwidth is doubled, and the L1 cache can now run as either 32K L1 and 64K shared memory, or 64K L1 and 32K shared memory. That potentially increases the L1 cache size 167 percent relative to Pascal. L2 cache size has also been doubled, and Nvidia states that the L2 cache delivers "significantly higher bandwidth" as well.
The improved clockspeeds of GDDR6 relative to GDDR5 and GDDR5X would help, but Turing doesn't stop there. We know Pascal already had several lossless memory compression techniques available, and the Turing architecture continues to improve in this area. Nvidia doesn't provide specific details on what has changed, but the larger caches and improved compression increase the effective bandwidth by 20-35 percent relative to Pascal GPUs.
Combined, the GeForce RTX 2080 Ti on average has 50 percent more effective bandwidth than the GTX 1080 Ti, even though memory speeds are only 27 percent faster. The RTX 2080 and 2070 should show even greater improvements, since the memory clocks have increased by 40 and 75 percent, respectively.
Turing architecture: even more enhancements
There's so much new stuff in Turing that it's hard to say how much any one aspect will matter in the long run. The Pascal and Maxwell architectures likewise had some new features—anyone remember VXAO, Voxel Ambient Occlusion, which to my knowledge was only used in two games (Rise of the Tomb Raider and Final Fantasy XV)? There's potential for these other features to be useful as well, but I'm grouping all of them together in the above gallery and this short description.
Mesh shaders are the next iteration of vertex, geometry, and tessellation shaders. The main idea here is to move LOD (Level of Detail) calculations from the CPU and onto the GPU. This can improve performance by orders of magnitude, and Nvidia showed a demonstration of a ship flying through a massive asteroid field with mesh shaders allowing for the real-time use of 'trillions' of polygons. The catch is that the LOD scaling culls that down to a more manageable number, in the millions instead of trillions. Mesh shaders will be an extension to existing graphics APIs for now, so they're less likely to see widespread use until/unless they're directly integrated into the DIrectX/Vulkan APIs, but the demo looked very cool.
Variable Rate Shading (VRS) is the next new feature, and it allows games to use more shaders where needed, and fewer shaders where it's not important. The goal is to provide equivalent image quality with better performance, and Nvidia suggested a 15 percent boost in performance should be possible. VRS can also be used in multiple ways, like MAS (Motion Adaptive Shading) where fast moving objects don't require as much detail (because they end up being blurred), and CAS (Content Adaptive Shading) where more effort is spent on complex surfaces like a car in a driving game, and less effort is used on simple surfaces like the road in the same game.
Nvidia showed a modified build of Wolfenstein II running with CAS and the ability to toggle the feature on/off. Without spending more time pixel peeping, I can say that there was no immediately visible difference between the two modes, but CAS did improve performance slightly. Whether we'll see a public patch to the game or not remains to be seen, and again this is something less likely to see widespread use.
Finally, Nvidia briefly discussed two more features in the Turing architecture: Multi-View Rendering (MVR), an enhanced version of the Simultaneous Multi-Projection (SMP) that was already a feature in Pascal, and Texture Space Shading (TSS). Where SMP primarily focused on two views and VR applications, MVP can do four views per pass and removes some view dependent attributes. It should help to further improve performance in VR applications, especially with some of the newer VR headsets that have a wider field of view.
TSS meanwhile makes even less sense to those of us not actively writing game engines. Nvidia says it can allow developers to exploit both spatial and temporal rendering redundancy, effectively reducing the amount of shader work that needs to run. There are several pages in the Turing architecture whitepaper describing use cases for TSS, but as with previous technologies like SMP and VXAO, it remains to be seen how many developers will use the feature.
Turing architecture: improved NVENC for videos
Outside of the GPU and memory enhancements, Nvidia has also worked to improve NVENC, the hardware used for video encoding/decoding. Pascal GPUs delivered good performance but the quality of the resulting videos wasn't always as good as even the x264 Fast profile running on a CPU. With Turing, Nvidia claims to deliver equal or better quality than x264 Fast, with almost no CPU load.
If you're streaming at 1080p, this won't matter much since Pascal could handle that resolution fine, and a good CPU could run x264 Fast encoding with only a modest overhead. Moving up to 4k and higher bitrates is a different matter, with CPU utilization spiking and a large number of dropped frames. Turing aims to deliver 4k encoding with almost no CPU impact.
Outside of streaming use, the Turing architecture also adds support for 8k30 HEVC HDR encoding, and can also deliver equivalent quality to Pascal at 15-25 percent lower bitrates for HEVC and H.264 content. For decoding, Turing adds support for VP9 10/12b HDR content and HEVC 444 10/12b HDR.
Turing architecture: manufactured using TSMC 12nm FinFET
The Turing architecture advancement are largely thanks to improvements in manufacturing technologies. Turing GPUs will be manufactured using TSMC's 12nm FinFET process. While TSMC 12nm FinFET is more of a refinement and tweak to the existing 16nm rather than a large reduction in feature sizes, optimizations to the process technology over the past two years should help improve clockspeeds, chip density, and power use—the holy trinity of faster, smaller, and cooler running chips. TSMC's 12nm FinFET process is also mature at this point, with good yields, which allows Nvidia to create such large GPUs.
Even with the process improvements, TU102, TU104, and TU106 are all very big. The TU104 as an example is only slightly smaller than the GP102 used in the GTX 1080 Ti. Further driving home the fact that 12nm is more marketing than an actual shrink, GP102 has 12 billion transistors compared to TU106's 10.8 billion transistors. That's 11 percent more transistors in the GP102, with a die size that's eight percent larger.
There's hope for future improvements sooner rather than later as well. TSMC is already close to full production for its 7nm process, and AMD's Vega 7nm Radeon Instinct GPUs are supposed to ship by the end of 2018. If TSMC 7nm works well, we could see a die shrink of Turing by late 2019. Maybe it will be called Ampere, maybe it will be something else. Not only would 7nm bring sizes down to more manageable levels, but Nvidia could double down on RT cores or other features.
Nvidia's Turing architecture is a game changer
With everything new in the Turing architecture, it's easy to see why Nvidia is calling this the biggest leap in graphics architectures the company has ever created. Real-time ray tracing or something similar has always been the pie in the sky dream of gamers. That dream just jumped 5-10 years closer.
Our graphics chips have come a long way in the past 30 years, including milestones like the 3dfx Voodoo as the first mainstream consumer card that could do high performance 3D graphics, the GeForce 256 as the first GPU with acceleration of the transform and lighting process, and AMD's Radeon 9700 Pro as the first fully programmable DirectX 9 GPU. Nvidia's Turing architecture looks to be as big of a change relative to its predecessors as any of those products.
Like all change, this isn't necessarily going to be a nice and clean break with the old and the beginning of something new. As cool as real-time ray-tracing might be, it requires new hardware. It's the proverbial chicken and egg problem, where the software won't support a new feature without the hardware, but building hardware to accelerate something that isn't currently used is a big investment. Nvidia has made that investment with RTX and its Turing architecture, and only time will tell if it pays off.
For at least the next five years, we're going to be in a messy situation where most gamers don't have a card that can do RTX or DXR ray tracing at acceptable performance levels—with or without machine learning Tensor cores to help with denoising. The good news is that DXR leverages many of the data structures already used in Direct3D, so adding some ray tracing effects to a game shouldn't be too difficult. That's good news, because most game developers will need to continue supporting legacy products and rasterization technologies. Hybrid rendering will be with us for a long time, I suspect.
As cool as ray tracing is, I also have to wonder if the machine learning capabilities in Turing may prove to be even more important. Deep learning is revolutionizing diverse fields including science, medicine, automotive, and more. I'm not convinced DLSS will really look as good as native 4k rendering, but then I can also say that the difference between 4k and 1440p in many games isn't nearly as great as you might expect. I'm cautiously optimistic, though, and even more so for things like improved AI, anti-cheat, and other potential uses.
The Turing architecture is a tour de force from the leading graphics technology company. It's far more than I expected to see when Volta was launched last year for supercomputers. It's also shocking how quickly the GV100 just lost most of its appeal. GeForce RTX 2080 Ti for $1,199 is a lot to ask, but it's also a helluva lot better deal than a Titan Xp for that same price, or a Titan V for $2,999.
But all the theoretical improvements for now don't make the GeForce RTX cards substantially faster than the existing Pascal solutions. In fact, the GTX 1080 Ti and RTX 2080 are effectively tied in most games, except the RTX card costs $100 more. The current situation brings to mind the transition from DirectX 7 (hardware transform and lighting) to DX8/DX9 (fully programmable shaders), where we had to wait years before the games properly utilized the new features. There are already more ray tracing games slated to launch in the next six months than early DX8 titles, but we're still left waiting for at least a month or two before we get to see the RT cores properly utilized in a game.
GPU microarchitecture by Nvidia
Turing is the codename for a graphics processing unit (GPU) microarchitecture developed by Nvidia. It is named after the prominent mathematician and computer scientist Alan Turing. The architecture was first introduced in August 2018 at SIGGRAPH 2018 in the workstation-oriented Quadro RTX cards, and one week later at Gamescom in consumer GeForce RTX 20 series graphics cards. Building on the preliminary work of its HPC-exclusive predecessor, the Turing architecture introduces the first consumer products capable of real-time ray tracing, a longstanding goal of the computer graphics industry. Key elements include dedicated artificial intelligence processors ("Tensor cores") and dedicated ray tracing processors. Turing leverages DXR, OptiX, and Vulkan for access to ray-tracing. In February 2019, Nvidia released the GeForce 16 series of GPUs, which utilizes the new Turing design but lacks the ray tracing and artificial intelligence cores.
Turing is manufactured using TSMC's 12 nmFinFETsemiconductor fabrication process. The high-end TU102 GPU includes 18.6 billion transistors fabricated using this process. Turing also uses GDDR6 memory from Samsung Electronics, and previously Micron Technology.
The Turing microarchitecture combines multiple types of specialized processor core, and enables an implementation of limited real-time ray tracing. This is accelerated by the use of new RT (ray-tracing) cores, which are designed to process quadtrees and spherical hierarchies, and speed up collision tests with individual triangles.
Features in Turing:
The GDDR6 memory is produced by Samsung Electronics for the Quadro RTX series. The RTX 20 series initially launched with Micron memory chips, before switching to Samsung chips by November 2018.
Nvidia reported rasterization (CUDA) performance gains for existing titles of approximately 30–50% over the previous generation.
The ray-tracing performed by the RT cores can be used to produce reflections, refractions and shadows, replacing traditional raster techniques such as cube maps and depth maps. Instead of replacing rasterization entirely, however, the information gathered from ray-tracing can be used to augment the shading with information that is much more photo-realistic, especially in regards to off-camera action. Nvidia said the ray-tracing performance increased about 8 times over the previous consumer architecture, Pascal.
Generation of the final image is further accelerated by the Tensor cores, which are used to fill in the blanks in a partially rendered image, a technique known as de-noising. The Tensor cores perform the result of deep learning to codify how to, for example, increase the resolution of images generated by a specific application or game. In the Tensor cores' primary usage, a problem to be solved is analyzed on a supercomputer, which is taught by example what results are desired, and the supercomputer determines a method to use to achieve those results, which is then done with the consumer's Tensor cores. These methods are delivered via driver updates to consumers. The supercomputer uses a large number of Tensor cores itself.
Main article: Nvidia RTX
Turing's development platform is called RTX. RTX ray-tracing features can be accessed using Microsoft's DXR, OptiX, as well using Vulkan extensions (the last one being also available on Linux drivers). It includes access to AI-accelerated features through NGX. The Mesh Shader, Shading Rate Image functionalities are accessible using DX12, Vulkan and OpenGL extensions on Windows and Linux platforms.
Windows 10 October 2018 update includes the public release of DirectX Raytracing.
Products using Turing
- GeForce 16 series
- GeForce GTX 1650
- GeForce GTX 1650 (Mobile)
- GeForce GTX 1650 Max-Q (Mobile)
- GeForce GTX 1650 (GDDR6)
- GeForce GTX 1650 Super
- GeForce GTX 1650 Ti (Mobile)
- GeForce GTX 1660
- GeForce GTX 1660 (Mobile)
- GeForce GTX 1660 Super
- GeForce GTX 1660 Ti
- GeForce GTX 1660 Ti (Mobile)
- GeForce GTX 1660 Ti Max-Q (Mobile)
- GeForce 20 series
- GeForce RTX 2060
- GeForce RTX 2060 (Mobile)
- GeForce RTX 2060 Max-Q (Mobile)
- GeForce RTX 2060 Super
- GeForce RTX 2060 Super (Mobile)
- GeForce RTX 2070
- GeForce RTX 2070 (Mobile)
- GeForce RTX 2070 Max-Q (Mobile)
- GeForce RTX 2070 Max-Q Refresh (Mobile)
- GeForce RTX 2070 Super
- GeForce RTX 2070 Super (Mobile)
- GeForce RTX 2070 Super Max-Q (Mobile)
- GeForce RTX 2080
- GeForce RTX 2080 (Mobile)
- GeForce RTX 2080 Max-Q (Mobile)
- GeForce RTX 2080 Super
- GeForce RTX 2080 Super (Mobile)
- GeForce RTX 2080 Super Max-Q (Mobile)
- GeForce RTX 2080 Ti
- Titan RTX
- Nvidia Quadro
- Quadro RTX 3000 (Mobile)
- Quadro RTX 4000
- Quadro RTX 5000
- Quadro RTX 6000
- Quadro RTX 8000
- Nvidia Tesla
- ^Tom Warren; James Vincent (May 14, 2020). "Nvidia's first Ampere GPU is designed for data centers and AI, not your PC". The Verge.
- ^"NVIDIA Announces the GeForce RTX 20 Series: RTX 2080 Ti & 2080 on Sept. 20th, RTX 2070 in October". Anandtech.
- ^"NVIDIA TURING GPU ARCHITECTURE: Graphics Reinvented"(PDF). Nvidia. 2018. Retrieved June 28, 2019.
- ^"Nvidia announces RTX 2000 GPU series with '6 times more performance' and ray-tracing". The Verge. Retrieved August 20, 2018.
- ^"The NVIDIA Turing GPU Architecture Deep Dive: Prelude to GeForce RTX". AnandTech.
- ^Mujtaba, Hassan (August 14, 2018). "Samsung GDDR6 Memory Powers NVIDIA's Turing GPU Based Quadro RTX Cards". wccftech.com. Retrieved June 19, 2019.
- ^Maislinger, Florian (November 21, 2018). "Faulty RTX 2080 Ti: Nvidia switches from Micron to Samsung for GDDR6 memory". PC Builder's Club. Retrieved July 15, 2019.
- ^ ab"#BeForTheGame". Twitch.tv.
- ^Jeff Fisher. "GeForce RTX Propels PC Gaming's Golden Age with Real-Time Ray Tracing". Nvidia.
- ^"NVIDIA RTX platform". Nvidia.
- ^"Turing Extensions for Vulkan and OpenGL". Nvidia.
Nvidia Turing Details: Specs, Price, DLSS and RTX Games
Catch up with all our Nvidia Turing reviews:
Nvidia Turing Release DateNvidia Turing possibly had the longest windup to its initial reveal at Gamescom 2018. Prior to that date, the entire PC gaming world was chomping at the bit for new graphics cards while the Bitcoin craze made GPUs a rare and expensive commodity, a period lasting between the last quarter of 2017 until the about the Spring of 2018.
Turing debuted with a trio of cards on August 20, 2018: the GeForce RTX 2080 Ti, GeForce RTX 2080, and GeForce RTX 2070. Although the three cards were announced together, they didn’t appear on shelves at the same time. The Nvidia GeForce RTX 2080 released on schedule on September 20, 2018, but the RTX 2080 Ti came a week later than planned on September 27, 2018. Meanwhile, the RTX 2070 didn’t release until October 17, 2018.
The next Turing graphics card wouldn’t be announced until CES 2019. The RTX 2060—and RTX-powered gaming laptops—were announced on January 7, 2019, and subsequently released on January 15, 2019.Following sluggish sales on RTX cards, a continued lack of ray-traced games, and plenty of blowback from the PC gaming community, Nvidia announced its three Turing GTX cards and released the Nvidia GeForce GTX 1660 Ti on February 22, 2019. The Nvidia GeForce GTX 1660 followed shortly after on March 14, 2019. Lastly, the GeForce GTX 1650 arrived on April 23, 2019, along with the first bundle of Turing GTX-powered gaming laptops.
Up until now, this has all been a slightly more frequent release cadence for Nvidia, but then Team Green announced Super GPUs on July 2, 2019, as a completely new addition for its lineup. The RTX 2070 Super and RTX 2060 Super both released on the same day of their announcement, and the RTX 2080 Super arrived a few weeks later on July 23, 2019.
Nvidia Turing PriceOverall with Turing, the price of Nvidia graphics cards has definitely increased since Pascal. What would have bought you an Nvidia GTX 1080 in the past will now buy you an Nvidia RTX 2070 Super, a GTX 1070 Ti for an RTX 2060 Super, and so on.
Essentially, Nvidia has set the goalposts forward with its pricing scheme on its RTX graphics cards. That said, the value you can squeeze out of a Turing GTX graphics card has never been better.
Here’s a quick breakdown of the launch pricing on all the Nvidia Turing GPUs:
- Nvidia GeForce RTX 2080 Ti: $1,199
- Nvidia GeForce RTX 2080 Super: $699
- Nvidia GeForce RTX 2080: $699
- Nvidia GeForce RTX 2070 Super: $499
- Nvidia GeForce RTX 2070: $499
- Nvidia GeForce RTX 2060 Super: $399
- Nvidia GeForce RTX 2060: $349
- Nvidia GeForce GTX 1660 Ti: $279
- Nvidia GeForce GTX 1660: $219
- Nvidia GeForce GTX 1650: $149
Nvidia Turing SpecsWith the introduction of Turing, Nvidia really did introduce a completely new GPU architecture. Not only did Turing make the jump from Pascal 14nm process to a 12nm one, but it also introduced a much larger die that added ray tracing and Tensor cores to the mix.
Ray tracing cores or RT cores are a specialized part of the GPU designed to essentially trace the rays of in-game light and calculate how they’re being reflected, refracted, and absorbed to produce more realistic graphics. Unlike most ray tracing used in computer graphics; RT cores are one of the few dedicated hardware solutions that do this work in real-time.
For now, RT cores remain to be an exclusive feature for Nvidia’s Turing RTX cards. However, less than a half year after the introduction Turing, Nvidia admitted ray tracing cores aren’t necessary to enable RTX mode. The GPU maker released a driver on April 11, 2019, allowing GTX graphics cards both new and old (down to the Nvidia GTX 1060) to take advantage of ray tracing.
Tensor cores, on the other hand, are designed for deep learning processes. Its most commonly used comes tied with Nvidia’s own Deep Learning Super-Sampling, which applies a specialized form of anti-aliasing that produces sharper images with higher framerates than other forms of AA can produce.
One other big component that Nvidia Turing introduced was GDDR6 memory, which can achieve speeds of up to (so far) 15.5Gbps. Comparatively, the maximum stock memory speed achieved on GDDR5 memory was 11.4GBps. Unfortunately, not all Turing graphics card feature GDDR6 memory as the GTX 1660 and GTX 1650 still only get GDDR5 memory.
New types of cores and video memory aside, the new Turing GPU itself is also fairly different from the Pascal GPU. For Turing, Nvidia streamlined its architecture by placing only a single warp scheduler and dispatch unit into each Streaming Multiprocessor (SM)—otherwise known as the smallest building blocks of a GPU.
This is also where Tensor Cores come into integrate itself directly with the Turing architecture to denoise images—or essentially fill in the blocks missing in rendered images. RT Cores, on the other hand, are almost completely divorced from the rest of the GPU’s main image processing block. In fact, these specialized RT Cores are only used to power Nvidia’s RTX technology or compute sound ray tracing. Otherwise, RT Cores remain completely inert, drawing zero power when not in use.
The new 16-series GTX gain all the streamlined improvements of Turing’s architecture, but not the specialized RT or Tensor Cores. In place of Tensor cores, Turing-based GTX cards have Floating Point 16 cores that handle integer heavy workloads.
But when it comes to real-time ray tracing, the only thing the GTX 1660 Ti, GTX 1660, and 1650 have to handle RTX is their own brute graphical strength—which has worked out surprisingly well so far.
Nvidia Turing RTX GamesDespite real-time ray tracing being Nvidia Turing RTX’s signature feature, the list of games that actually take advantage of the technology is still fairly short. That said, the list has been growing and will likely explode when the PS5 and Xbox Scarlett hit the market, which are said to support ray tracing and may release next year.
Here are the games we know that have, or will, support RTX, when they enabled their respective ray tracing features, and the type of ray tracing you can expect from them (which we've denoted as “RTX Effects” whenever the developer or Nvidia hasn’t specified what the “effects” exactly are):
- Assetto Corsa (Sept 2018): RTX Reflections
- Battlefield V (Nov 2018): RTX Reflections
- JX3 (Nov 2018): “RTX Effects”
- Metro Exodus (Feb 2019): RTX Global Illumination
- Shadow of the Tomb Raider (March 2019): RTX Procedural Lighting and Shadows
- Stay in the Light (June 2019): RTX Shadows and Lighting
- Quake II RTX (June 2019): RTX Global Illumination
- Wolfenstein: Youngblood (July 2019): “RTX Effects”
- Control (Aug 2019): RTX Reflections, Transparent Reflections, Diffuse Global Illumination, and Contact Shadows
- Mechwarrior V: Mercenaries (Sept 2019): RTX Reflections
- Call of Duty: Modern Warfare (Oct 2019): “RTX Effects”
- Atomic Heart (Q4 2019): RTX Ambient Occlusion, Reflections, and Shadows
- Enlisted (TBD 2019): RTX Global Illumination
- Vampire: The Masquerade – Bloodlines 2 (Mar 2020): “RTX Effects”
- CyberPunk 2077 (Apr 2020): “RTX Effects”
- Watch Dogs: Legion (Mar 2020): “RTX Effects”
- Justice (TBD): RTX Reflections and Shadows
- Dragon Hound (TBD): “RTX Effects”
Nvidia Turing DLSS GamesWhile Nvidia’s RTX implementation has been varied with what types of ray tracing users can expect, DLSS has been much cleaner cut.
For the most part, every game on this list below launched with or will receive the ability to utilize Team Green’s AI-based anti-aliasing technology on these following dates (unfortunately for many, DLSS support has been only announced):
- JX3 – November 20, 2018
- Final Fantasy 15 – December 12, 2018
- Metro Exodus – February 13, 2019
- Battlefield V – February 13, 2019
- Shadow of the Tomb Raider – March 20, 2019
- Anthem – March 26, 2019
- Monster Hunter World – July 17, 2019
- Remnant: From The Ashes – August 20, 2019
- Mechwarrior 5: Mercenaries – September 10, 2019
- SCUM – November 9, 2019
- Stormdivers – November 9, 2019
- Vampire: The Masquerade – Bloodlines 2 – March 2020
- Hitman 2 – Announced August 2018, Arrival TBD
- Islands of Nyne – Announced August 2018, Arrival TBD
- Justice – Announced August 2018, Arrival TBD
- KINETIK – Announced September 2018, Arrival TBD
- Fractured Lands – Announced September 2018, Arrival TBD
- Ark: Survival Evolved – Announced September 2018, Arrival TBD
- Dauntless – Announced September 2018, Arrival TBD
- Deliver Us The Moon: Fortuna – Announced September 2018, Arrival TBD
- Fear the Wolves – Announced September 2018, Arrival TBD
- Hellblade: Senua’s Sacrifice – Announced September 2018, Arrival TBD
- Outpost Zero – Announced September 2018, Arrival TBD
- Overkill’s The Walking Dead – Announced September 2018, Arrival TBD
- PlayerUnknown’s Battlegrounds – Announced September 2018, Arrival TBD
- Serious Sam 4: Planet Badass – Announced September 2018, Arrival TBD
- The Forge Arena – Announced September 2018, Arrival TBD
- We Happy Few – TBD
- Atomic Heart – TBD
- Darksiders III – TBD
Was this article informative?
GeForce 20 series
Series of GPUs by Nvidia
The GeForce 20 series is a family of graphics processing units developed by Nvidia. Serving as the successor to the GeForce 10 series, the line started shipping on September 20, 2018, and after several editions, on July 2, 2019, the GeForce RTX Super line of cards was announced.
The 20 series marked the introduction of Nvidia's Turing microarchitecture, and the first generation of RTX cards, the first in the industry to implement realtimehardwareray tracing in a consumer product. In a departure from Nvidia's usual strategy, the 20 series doesn't have an entry level range, leaving it to the 16 series to cover this segment of the market.
These cards are succeeded by the GeForce 30 series, powered by the Ampere microarchitecture.
On August 14, 2018, Nvidia teased the announcement of the first card in the 20 series, the GeForce RTX 2080, shortly after introducing the Turing architecture at SIGGRAPH earlier that year. The GeForce 20 series was finally announced at Gamescom on August 20, 2018, becoming the first line of graphics cards "designed to handle real-time ray tracing" thanks to the "inclusion of dedicated tensor and RT cores."
In August 2018, it was reported that Nvidia had trademarked GeForce RTX and Quadro RTX as names.
The line started shipping on September 20, 2018. Serving as the successor to the GeForce 10 series, the 20 series marked the introduction of Nvidia's Turing microarchitecture, and the first generation of RTX cards, the first in the industry to implement realtimehardwareray tracing in a consumer product.
Released in late 2018, the RTX 2080 was marketed as up to 75% faster than the GTX 1080 in various games, also describing the chip as "the most significant generational upgrade to its GPUs since the first CUDA cores in 2006," according to PC Gamer.
After the initial release, factory overclocked versions were released in the fall of 2018. The first was the "Ti" edition, while the Founders Edition cards were overclocked by default and had a three-year warranty. When the GeForce RTX 2080 Ti came out, TechRadar called it "the world’s most powerful GPU on the market." The GeForce RTX 2080 Founders Edition was positively reviewed for performance by PC Gamer on September 19, 2018, but was criticized for the high cost to consumers, also noting that its ray tracing feature wasn't yet utilized by many programs or games. In January 2019, Tom's Hardware also stated the GeForce RTX 2080 Ti Xtreme was "the fastest gaming graphics card available," although it criticized the loudness of the cooling solution, the size and heat output in PC cases. In August 2018, the company claimed that the GeForce RTX graphics cards were the "world’s first graphics cards to feature super-fast GDDR6 memory, a new DisplayPort 1.4 output that can drive up to 8K HDR at 60Hz on future-generation monitors with just a single cable, and a USB Type-C output for next-generation Virtual Reality headsets."
In October 2018, PC Gamer reported the supply of the 2080 Ti card was "extremely tight" after availability had already been delayed. By November 2018, MSI was offering nine different RTX 2080-based graphics cards. Released in December 2018, the line's Titan RTX was initially priced at $2500, significantly more than the $1300 then needed for a GeForce RTX 2080 Ti.
In January 2019, Nvidia announced that GeForce RTX graphics cards would be used in 40 new laptops from various companies. Also that month, in response to negative reactions to the pricing of the GeForce RTX cards, Nvidia CEO Jensen Huang stated "They were right. [We] were anxious to get RTX in the mainstream market... We just weren’t ready. Now we’re ready, and it’s called 2060," in reference to the RTX 2060. In May 2019, a TechSpot review noted that the newly released Radeon VII by AMD was comparable in speeds to the GeForce RTX 2080, if slightly slower in games, with both priced similarly and framed as direct competitors.
On July 2, 2019, the GeForce RTX Super line of cards was announced, which comprises higher-spec versions of the 2060, 2070 and 2080. Each of the Super models were offered for a similar price as older models but with improved specs. In July 2019, NVidia stated the "SUPER" graphics cards in the GeForce RTX 20 series, to be introduced, had a 15% performance advantage over the GeForce RTX 2060.PC World called the super editions a "modest" upgrade for the price, and the 2080 Super chip the "second most-powerful GPU ever released" in terms of speed. In November 2019, PC Gamer wrote "even without an overclock, the 2080 Ti is the best graphics card for gaming." In June 2020, PC Mag listed the Nvidia GeForce RTX 2070 Super as one of the "best  graphics cards for 4k gaming in 2020." The GeForce RTX 2080 Founders Edition, Super, and Ti were also listed. In June 2020, graphic cards including the RTX 2060, RTX 2060 Super, RTX 2070 and the RTX 2080 Super were announced as discounted by retailers in expectation of the GeForce RTX 3080 launch. In April 2020, Nvidia announced 100 new laptops licensed to include either GeForce GTX and RTX models.
Reintroduction of older cards
Due to production problems surrounding the RTX 30-series cards and a general shortage of graphics cards due to production issues caused by the ongoing COVID-19 pandemic, which led to a global shortage of semiconductor chips, and general demand for graphics cards increasing due to an increase in cryptocurrency mining, the RTX 2060 and its Super counterpart, alongside the GTX 1050 Ti, were brought back into production in 2021.
See also: Turing (microarchitecture) and Ray-tracing hardware
The RTX 20 series is based on the Turing microarchitecture and features real-time hardware ray tracing. The cards are manufactured on an optimized 14 nm node from TSMC, named 12 nm FinFET NVIDIA (FFN). New example features in Turing included mesh shaders, rRay tracing (RT) cores (bounding volume hierarchy acceleration), tensor (AI) cores, dedicated Integer (INT) cores for concurrent execution of integer, and floating point operations. In the GeForce 20 series, this real-time ray tracing is accelerated by the use of new RT cores, which are designed to process quadtrees and spherical hierarchies, and speed up collision tests with individual triangles.
The ray tracing performed by the RT cores can be used to produce effects such as reflections, refractions, shadows, depth of field, light scattering and caustics, replacing traditional raster techniques such as cube maps and depth maps. Notes: Instead of replacing rasterization entirely, however, ray tracing is offered in a hybrid model, in which the information gathered from ray tracing can be used to augment the rasterized shading for more photo-realistic results.
The second generation Tensor Cores (succeeding Volta's) work in cooperation with the RT cores, and their AI features are used mainly to two ends: firstly, de-noising a partially ray traced image by filling in the blanks between rays cast; also another application of the Tensor cores is DLSS (deep learning super-sampling), a new method to replace anti-aliasing, by artificially generating detail to upscale the rendered image into a higher resolution. The Tensor cores apply deep learning models (for example, an image resolution enhancement model) which are constructed using supercomputers. The problem to be solved is analyzed on the supercomputer, which is taught by example what results are desired. The supercomputer then outputs a model which is then executed on the consumer's Tensor cores. These methods are delivered to consumers as part of the cards' drivers.
Nvidia segregates the GPU dies for Turing into A and non-A variants, which is appended or excluded on the hundreds part of the GPU code name. Non-A variants are not allowed to be factory overclocked, whilst A variants are.
The GeForce 20 series was launched with GDDR6 memory chips from Micron Technology. However, due to reported faults with launch models, Nvidia switched to using GDDR6 memory chips from Samsung Electronics by November 2018.
Main article: Nvidia RTX
With the GeForce 20 series, Nvidia introduced the RTX development platform. RTX uses Microsoft's DXR, Nvidia's OptiX, and Vulkan for access to ray tracing. The ray tracing technology used in the RTX Turing GPUs was in development at Nvidia for 10 years. Nvidia's Nsight Visual Studio Edition application is used to inspect the state of the GPUs.
All of the cards in the series are PCIe 3.0 x16 cards, manufactured using a 12 nmFinFET process from TSMC, and use GDDR6 memory (initially Micron chips upon launch, and later Samsung chips from November 2018).
|Model||Launch||Code name(s)||Transistors (billion)||Die size|
|Shader processors||Texture mapping units||Render output units||Ray tracing cores||Tensor cores[a]||SM|
|Clock speeds||Fillrate||Memory||Processing power (GFLOPS)||Ray tracing performance||TDP|
|Launch MSRP (USD)||Code name(s)||Model|
|Base core clock|
|Boost core clock|
|GeForce RTX 2060||January 15, 2019||TU106-200A-KA-A1||10.8||445||1920||120||48||30||240||30||3||1365||1680||14000||65.52||163.8||6||336||192||5242 (6451)||164 (202)||10483 (12902)||5||37||51.6||160||No||$349||TU106-200A-KA-A1||GeForce RTX 2060|
|GeForce RTX 2060 TU104||January 10, 2020||TU104-150-KC-A1||13.6||545||$300||TU104-150-KC-A1||GeForce RTX 2060 TU104|
|GeForce RTX 2060 Super||July 9, 2019||TU106-410-A1||10.8||445||2176||136||64||34||272||34||4||1470||1650||94.05||199.9||8||448||256||6123 (7181)||191 (224)||12246 (14362)||6||41||57.4||175||$399||TU106-410-A1||GeForce RTX 2060 Super|
|GeForce RTX 2070||October 17, 2018||TU106-400-A1||2304||144||36||288||36||1410||1620||90.24||203.04||6497 (7465)||203 (233)||12994 (14930)||45||59.7||$499||N/A||TU106-400-A1||GeForce RTX 2070|
|TU106-400A-A1||1620+||6497 (7465+)||203 (233+)||12994 (14930+)||$499+||$599||TU106-400A-A1|
|GeForce RTX 2070 Super||July 9, 2019||TU104-410-A1||13.6||545||2560||160||40||320||40||1605||1770||102.72||256.8||8218 (9062)||257 (283)||16435 (18125)||7||52||72.5||215||2-way NVLink||$499||TU104-410-A1||GeForce RTX 2070 Super|
|GeForce RTX 2080||September 20, 2018||TU104-400-A1||2944||184||46||368||46||1515||1710||96.96||278.76||8920 (10068)||279 (315)||17840 (20137)||8||60||80.5||$699||N/A||TU104-400-A1||GeForce RTX 2080|
|TU104-400A-A1||1710+||8920 (10068+)||279 (315+)||17840 (20137+)||$699+||$799||TU104-400A-A1|
|GeForce RTX 2080 Super||July 23, 2019||TU104-450-A1||3072||192||48||384||48||1650||1815||15500||105.6||316.8||496||10138 (11151)||317 (349)||20275 (22303)||63||89.2||250||$699||TU104-450-A1||GeForce RTX 2080 Super|
|GeForce RTX 2080 Ti||September 27, 2018||TU102-300-K1-A1||18.6||754||4352||272||88||68||544||68||5.5||1350||1545||14000||118.8||367.2||11||616||352||11750 (13448)||367 (421)||23500 (26896)||10||78||107.6||$999||N/A||TU102-300-K1-A1||GeForce RTX 2080 Ti|
|TU102-300A-K1-A1||1545+||11750 (13448+)||367 (421+)||23500 (26896+)||$999+||$1,199||TU102-300A-K1-A1|
|Nvidia Titan RTX||December 18, 2018||TU102-400-A1||4608||288||96||72||576||72||6||1770||129.6||388.8||24||672||384||12442 (16312)||389 (510)||24884 (32625)||11||84||130.5||280||$2,499||TU102-400-A1||Nvidia Titan RTX|
- ^A Tensor core is a mixed-precision FPU specifically designed for matrix arithmetic.
- ^The number of Streaming multi-processors on the GPU.
- ^Pixel fillrate is calculated as the lowest of three numbers: number of ROPs multiplied by the base core clock speed, number of rasterizers multiplied by the number of fragments they can generate per rasterizer multiplied by the base core clock speed, and the number of streaming multiprocessors multiplied by the number of fragments per clock that they can output multiplied by the base clock rate.
- ^Texture fillrate is calculated as the number of TMUs multiplied by the base core clock speed.
- ^"Introducing NVIDIA GeForce RTX 2070 Graphics Card". NVIDIA. Retrieved August 20, 2018.
- ^"NVIDIA GeForce RTX 2080 Founders Edition Graphics Card". NVIDIA. Retrieved August 20, 2018.
- ^"Graphics Reinvented: NVIDIA GeForce RTX 2080 Ti Graphics Card". NVIDIA. Retrieved August 20, 2018.
- ^ ab"GeForce RTX 2080 launch live blog: Nvidia's Gamescom press conference as it happens". TechRadar. Retrieved August 21, 2018.
- ^ abSamit Sarkar. "Nvidia unveils powerful new RTX 2070, RTX 2080, RTX 2080 Ti graphics cards". Polygon. Retrieved August 20, 2018.
- ^ ab"Nvidia's new RTX 2080, 2080 Ti video cards shipped on Sept 20, 2018, starting at $799". Ars Technica. Retrieved August 20, 2018.
- ^ abLori Grunin (July 2, 2019). "Nvidia's GeForce RTX Super line boosts 2060, 2070 and 2080 for same $$". CNET. Retrieved July 16, 2020.
- ^ abChuong Nguyen (August 14, 2018). "Nvidia teases new GeForce RTX 2080 launch at Gamescom next week". Digital Trends. Retrieved July 16, 2020.
- ^ abcBrad Chacos (September 19, 2018). "Nvidia Turing GPU deep dive: What's inside the radical GeForce RTX 2080 Ti". PCWorld. Retrieved July 16, 2020.
- ^"NVIDIA GeForce GTX 16 Series Graphics Card". NVIDIA. Retrieved October 31, 2020.
- ^Kevin Lee (August 10, 2018). "GeForce RTX 2080 may be the name of Nvidia's next flagship graphics card". Tech Radar. Retrieved July 21, 2020.
- ^ abTom Warren and Stefan Etienne (September 19, 2018). "Nvidia GeForce RTX 2080 Review: 4k Gaming is Here, At a Price". The Verge. Retrieved July 16, 2020.
- ^Jarred Walton (October 8, 2018). "Nvidia GeForce RTX 2080: benchmark, release date, and everything you need to know". PC Gamer. Retrieved July 21, 2020.
- ^Gabe Carey (November 21, 2018). "PNY GeForce RTX 2080 XLR8 Gaming Overclocked Edition Review". PC Mag. Retrieved July 21, 2020.
- ^Brad Chacos (August 25, 2018). "Nvidia's GeForce RTX 2080 and RTX 2080 Ti are loaded with boundary-pushing graphics tech". PCWorld. Retrieved July 16, 2020.
- ^Kevin Lee (November 15, 2019). "Nvidia GeForce RTX 2080 Ti review". Tech Radar. Retrieved July 16, 2020.
- ^ abcJarred Walton (September 19, 2018). "NVidia GEForce RTX 2080 Founders Edition Review". PC Gamer. Retrieved July 16, 2020.
- ^Chris Angelini, Igor Wallossek (September 19, 2018). "Nvidia GeForce RTX 2080 Founders Edition Review: Faster, More Expensive Than GeForce GTX 1080 Ti". Tom's Hardware. Retrieved July 16, 2020.
- ^Chris Angelini (January 1, 2019). "Aorus GeForce RTX 2080 Ti Xtreme 11G Review: In A League of its Own". Tom's Hardware. Retrieved July 21, 2020.
- ^Andrew Burnes (August 20, 2018). "GeForce RTX Founders Edition Graphics Cards: Cool and Quiet, and Factory Overclocked". www.nvidia.com. Nvidia. Retrieved August 1, 2020.
- ^Paul Lilly (October 30, 2018). "Some users are complaining of GeForce RTX 2080 Ti cards dying". PC Gamer. Retrieved July 21, 2020.
- ^Charles Jefferies (November 16, 2018). "MSI GeForce RTX 2080 Gaming X Trio Review". PC Mag. Retrieved July 21, 2020.
- ^Antony Leather (December 4, 2018). "Nvidia's Monster Titan RTX Has $2500 Price Tag". Forbes. Retrieved July 21, 2020.
- ^Andrew Burnes (January 6, 2019). "GeForce RTX GPUs Come to 40+ Laptops, Global Availability January 29". nvidia.com. NVidia. Retrieved August 1, 2020.
- ^Gordon Mah Ung (January 9, 2019). "Nvidia disses the Radeon VII, vowing the RTX 2080 will crush AMD's 'underwhelming' GPU". PCWorld. Retrieved July 21, 2020.
- ^Steven Walton (May 22, 2019). "Radeon VII vs. GeForce RTX 2080". TechSpot. Retrieved July 21, 2020.
- ^Andrew Burnes (July 2, 2019). "Introducing GeForce RTX SUPER Graphics Cards: Best In Class Performance, Plus Ray Tracing". www.nvidia.com. GeForce Nvidia. Retrieved August 1, 2020.
- ^Brad Chacos (July 23, 2019). "Nvidia GeForce RTX 2080 Super Founders Edition review: A modest upgrade to a powerful GPU". PCWorld. Retrieved July 16, 2020.
- ^Paul Lilly (November 4, 2019). "This external graphics box contains a liquid-cooled GeForce RTX 2080 Ti". PC Gamer. Retrieved July 21, 2020.
- ^John Burek and Chris Stobing (June 6, 2020). "The Best Graphics Cards for 4K Gaming in 2020". PC Mag. Retrieved July 21, 2020.
- ^Matt Hanson (June 29, 2020). "Nvidia graphics cards are getting price cuts ahead of expected RTX 3080 launch". Tech Radar. Retrieved July 16, 2020.
- ^"Announcing New GeForce Laptops, Combining New Max-Q Tech with GeForce RTX SUPER GPUs, For Up To 2X More Efficiency Than Last-Gen". nvidia.com. Nvidia. April 2, 2020. Retrieved August 1, 2020.
- ^Tom Warren (August 20, 2018). "Nvidia announces RTX 2000 GPU series with '6 times more performance' and ray-tracing". The Verge. Retrieved August 20, 2018.
- ^"NVIDIA Announces the GeForce RTX 20 Series: RTX 2080 Ti & 2080 on Sept. 20th, RTX 2070 in October". Anandtech. August 20, 2018. Retrieved December 6, 2018.
- ^Christoph Kubisch (September 17, 2018). "Introduction to Turing Mesh Shaders". Retrieved September 1, 2019.
- ^Nate Oh (September 14, 2018). "The NVIDIA Turing GPU Architecture Deep Dive: Prelude to GeForce RTX". AnandTech.
- ^Ryan Smith (August 13, 2018). "NVIDIA Reveals Next-Gen Turing GPU Architecture: NVIDIA Doubles-Down on Ray Tracing, GDDR6, & More". AnandTech.
- ^"NVIDIA Deep Learning Super-Sampling (DLSS) Shown To Press". www.legitreviews.com. August 22, 2018. Retrieved September 14, 2018.
- ^"NVIDIA Segregates Turing GPUs; Factory Overclocking Forbidden on the Cheaper Variant". TechPowerUP. September 17, 2018. Retrieved December 7, 2018.
- ^ abMaislinger, Florian (November 21, 2018). "Faulty RTX 2080 Ti: Nvidia switches from Micron to Samsung for GDDR6 memory". PC Builder's Club. Retrieved July 15, 2019.
- ^Florian Maislinger (November 21, 2018). "NVIDIA RTX platform". Nvidia.
- ^NVIDIA GeForce (August 20, 2018). "GeForce RTX - Graphics Reinvented". Youtube.
- ^"NVIDIA Nsight Visual Studio Edition". developer.nvidia.com. NVidia.
- ^ abNVIDIA no longer differentiates A and non-A GeForce RTX 2070 and 2080 dies after May 2019, with later dies for the affected models marked without 'A' suffix. "Nvidia to Stop Binning Turing A-Dies For GeForce RTX 2080 And RTX 2070 GPUs: Report". Tom's Hardware.
- ^ ab"NVIDIA GeForce RTX 2060 Graphics Card". NVIDIA.
- ^ abcdefSmith, Ryan. "The GeForce RTX 2070 Super & RTX 2060 Super Review: Smaller Numbers, Bigger Performance". www.anandtech.com. Retrieved July 3, 2019.
- ^ abcdef"Your Graphics, Now With SUPER Powers". NVIDIA. Retrieved July 3, 2019.
- ^ ab"NVIDIA GeForce RTX 2070 Graphics Card". NVIDIA.
- ^ ab"NVIDIA GeForce RTX 2080 Founders Edition Graphics Card". NVIDIA.
- ^ ab"Graphics Reinvented: NVIDIA GeForce RTX 2080 Ti Graphics Card". NVIDIA.
- ^ ab"NVIDIA TITAN RTX". NVIDIA. Retrieved December 18, 2018.
Nvidia Turing release date, news and features
The verdicts are in
Nvidia GeForce RTX 2080 Ti:
5 stars | High fps 4K gaming on one card; Leads ray tracing revolution in gaming | Extremely expensive; few initial ray tracing-supported games
Nvidia GeForce RTX 2080 Super:
4 stars | Great 1440p and 4K gaming performance; Cheaper than original RTX 2080; FrameView software is useful | Still expensive; Minimal performance gains over RTX 2080
Nvidia GeForce RTX 2080:
4.5 stars | Impressively improved gaming performance; Super simple overclocking | Nvidia’s most expensive xx80 card yet; More power demanding
Nvidia GeForce RTX 2070 Super:
5 stars | Founders Edition cheaper than original 2070, More CUDA cores, 1440p gaming with ray tracing | Still kind of expensive, Founders Edition card is heavy
Nvidia GeForce RTX 2070:
4 stars | Playable 4K gaming; Impressive synthetic performance | Expensive for a mid-range GPU; No SLI
Nvidia GeForce RTX 2060 Super:
4.5 stars | Excellent 1440p performance, Shiny new card design, Affordable | Can't handle 4K gaming, Founders Edition is heavy
Nvidia GeForce RTX 2060:
5 stars | Silky ray traced 1080p gaming; Runs cooler than previous generation | Slightly pricier than predecessor; RTX heavily impacts performance at QHD and 4K
While it seemed to take forever for Nvidia to release its next-generation Turing graphics cards, it was worth the wait once they arrived. This architecture has been satisfying gamer’s graphics needs for some time now and the impressive performance out of these GPUs has been duly noted by the computer world.
The Nvidia Turing lineup is quite striking, from the original Nvidia GeForce RTX 2080 Ti, RTX 2080, RTX 2070 and RTX 2060, to the newer Super RTX cards led by the RTX 2080 Super. They’re still among the best graphics cards available to power the next couple years of gaming (especially considering how hard it is to get one of the newer generation Nvidia Ampere GPUs). And, don’t forget the Titan RTX card for the prosumer crowd.
Real-time ray tracing has come to the masses thanks to Nvidia Turing, revolutionizing how GPUs render the best PC games. And, considering how games like Metro Exodus and Shadow of the Tomb Raider look with ray tracing turned on, there’s no turning back. You can even enable ray tracing on non-RTX cards (though you will suffer a major performance hit).
And, with these graphics cards showcasing their graphics prowess in the desktop arena, it’s no surprise that Nvidia is hitting the mobile sphere as well. Its Turing-basedGeForce RTX 2080 Super and RTX 2070 Super graphics cards are now available for gaming laptops. You can now find a ton of laptops equipped with Nvidia Turing’s mobile GPUs.
Of course, there are Nvidia Turing cards out there for people that don’t need as much power: the GeForce GTX 1660 Ti, GeForce GTX 1660 and GeForce GTX 1650. They’re not quite as powerful as the RTX 2000 cards, but they’re significantly more affordable. On the other side of the spectrum, there’s the Nvidia Ampere series, Turing’s successor, which are blowing the competition away.
Still, the Nvidia Turing cards remain among the best on the market today, and the most influential graphics cards to ever hit the shelves.
Cut to the chase
- What is it? Nvidia’s latest graphics card architecture
- When is it out? Out since September 20
- What will it cost? $599 (£569, AU$899) - $10,000 (£7,830, AU$13,751)
Nvidia Turing release date
All of the currently-announced Nvidia Turing GPUs are now out in the wild – from the GTX 1660 and 1660 Ti to the RTX Super cards: RTX 2060 Super, RTX 2070 Super and RTX 2080 Super.
At CES 2019, we didn’t just finally get an RTX 2060 announcement, but also over 40 gaming laptops sporting the mobile version of Nvidia RTX graphics. These days, the best gaming laptops, alongside the ones we saw at CES 2019 like the Alienware Area 51m, are all packing the latest Nvidia Turing graphics.
On the other hand, the GeForce RTX 2060 Super and GeForce RTX 2070 Super hit the streets in July to compete with AMD Navi. If you were looking for a graphics card that you won’t have to take out a personal loan to afford, these are certainly the more affordable options. The pricier GeForce RTX 2080 Super followed, seeing a release date of July 23rd.
Thankfully, the RTX-series cards are readily available now after some initial limited availability. And, thanks to the release of the newer cards, you’ll likely find more than a few older models on sale every day.
What’s more is users will have these RTX Super cards available on-the-go as well. Nvidia just recently announced that the RTX 2080 Super and RTX 2070 Super graphics cards are now going to be available for gaming laptops in April 2020. That’s alongside the GeForce RTX 2060 laptops, which are going to start coming in at $999 (about £800, AU$1,640).
Nvidia Turing price
Although the Nvidia Turing series started with the Quadro RTX GPUs, we're far more interested in the graphics cards available for consumers. If you wanted to check out these enterprise-leaning parts, we’ve got you covered here. Otherwise, the prices for Nvidia Turing graphics cards are as follows:
- Nvidia GeForce RTX 2080 Ti: $1,199 (£1,099, AU$1,899)
- Nvidia GeForce RTX 2080 Super: $699 (about £560, AU$1099)
- Nvidia GeForce RTX 2080: $799 (£749, AU$1,199)
- Nvidia GeForce RTX 2070 Super: $499 (about £395, AU$720)
- Nvidia GeForce RTX 2070: $599 (£569, AU$899)
- Nvidia GeForce RTX 2060 Super: $399 (about £315, AU$580)
- Nvidia GeForce RTX 2060: $349 (£329, AU$599)
- Nvidia GeForce GTX 1660 Ti: $279 (£259, AU$469)
- Nvidia GeForce GTX 1660: $219 (£219, AU$389)
- Nvidia GeForce GTX 1650: $149 (about £115, AU$210)
Overall, the prices for Nvidia's newest graphics cards seem to have risen with the Nvidia GeForce RTX 2080 Ti taking the place of Nvidia’s past Titan cards. This shift up was sadly seen across the entire lineup.
That is, until the Super RTX cards rolled out. Perhaps in an effort to keep up with the growing popularity of AMD’s Radeon RX 5700 GPUs, Nvidia released the RTX 2070 Super and RTX 2080 Super, at a lower price – $100 less, to be exact – than their predecessors when they were first released. Essentially, anyone looking to upgrade to one of the RTX Turing cards can now pay less for even better performance.
Then there’s the Turing GTX cards. These are heralded by the $279 (£259, AU$469) GTX 1660 Ti, and provide phenomenal value at the low end. The most recent of these cards, the GTX 1650, is priced at $149 (about £115, AU$210), and is positioned to compete with the AMD Radeon RX 570.
Nvidia Turing specs
The marquee feature of Nvidia Turing is the inclusion of ray-tracing, which can render more realistic visuals and lighting in real time without having to fall back on programming tricks. Through specialized RT cores, these graphics cards can basically calculate how light and sound travel in a 3D environment at a rate of up to 10 GigaRay/s on the RTX 2080 Ti. These specialized cores also allow Nvidia Turing-based graphics cards to process ray tracing up to 25 times faster than Pascal.
When these RT Cores aren’t in use for ray tracing, they essentially switch off, ceasing to draw any power.
In addition to these RT cores, the Turing Architecture also features Tensor Cores, like the ones found in Volta. These specialized cores enable artificial intelligence and neural networking so that Turing cards get better at rendering over time – something previously exclusive to supercomputers.
With the ability to deliver 500 trillion Tensor operations a second, this technology accelerates deep learning training and inferencing. This allows Nvidia to offer Deep Learning Super Sampling (DLSS), which could be a version of super sampling that won’t bring your computer to its knees.
Even for games that don’t support this new DLSS tech, these AI-fueled cores should deliver traditional anti-aliasing much more efficiently – up to eight times.
As with Volta, Nvidia Turing is adopting GDDR6 memory – up to 11GB in the RTX 2080 Ti, which can clock in at up to 14Gbps, and up to 8GB in the RTX 2080 Super, clocking in at up to 15.5Gbps. That’s quite the leap over the Pascal-powered Nvidia Titan Xp that clocked in at 11.4Gbps.
The Nvidia GeForce RTX 2080 Ti is an absolute behemoth of a GPU. With 4,352 CUDA cores, 11GB of GDDR6 VRAM with a 352-bit memory bus and 18 billion transistors, it’s capable of 4K Ultra gaming at high refresh rates for years to come. It’s no wonder it comes with such a high price tag.
The more mainstream RTX 2080 and RTX 2070, as well as the RTX 2060 Super and RTX 2070 Super, are also quite impressive and absolutely destroy the previous generation of GPUs. The former will feature 2,944 CUDA cores, 8GB of GDDR6 memory and clocks in at a 1.5GHz base frequency. The 2070, though is be a bit weaker, coming with 2,304 CUDA cores 8GB of GDDR6 VRAM and clocked at a 1,410Mhz base frequency.
And, while the RTX 2060 is basically just a cut-down RTX 2070, with the same TU106 GPU, but with 1,920 CUDA cores, 6GB of GDDR6 VRAM and a boost clock of 1,680 MHz, it’s still a formidable graphics card.
Nvidia has also launched some non-RTX cards, starting with the GTX 1660 Ti. This entry-level card features 1,536 CUDA cores, 6GB of GDDR6 VRAM at 12Gbps, and a base clock of 1,500 MHz. It’s slower than the RTX 2060, but it’s a substantial upgrade over the GTX 1060 it replaces.
Team Green’s second non-RTX GPU, the GTX 1660 features 1,480 CUDA cores, 6GB of GDDR5 video memory and a reference boost clock of 1,785MHz. It might not sound mighty on paper, but between its low price and fantastic 1080p gaming performance, it’s currently the absolutely best entry-level graphics card you can buy.
A third non-RTX GPU has been released, the Nvidia GeForce GTX 1650. This low-end GPU features the TU117 GPU, clocked at 1,485GHz with a boost of 1,665GHz. This budget card features 4GB of GDDR5 VRAM with 128GB/s of memory bandwidth on a 128-bit bus.
Nvidia Turing Performance
As long as you have the high-end specs to back them up, the new Turing RTX cards are able to perform much faster than their Pascal equivalents. These cards will be able to push it even further once DLSS or deep learning super sampling is more widespread. Additionally, due to the AA improvements in the Tensor cores, we’re seeing about a 20-40% increase in games that don't support DLSS.
In our benchmarks, the GeForce RTX 2080 outperforms the GeForce GTX 1080 Ti by about 11% and the Nvidia GTX 1080 by a more impressive 32% in Middle Earth: Shadow of War in 4K. This performance difference is even more massive when you look at the Nvidia GeForce RTX 2080 Ti. It’s not only 20% faster than the RTX 2080 in the same title, but beats out the last-generation 1080 Ti by a massive 30%, destroying the GTX 1080 with a 45% performance delta.
Then, there’s the RTX 2080 Super, the high-end offering in the Super RTX series. When pitted against its predecessor, we found the RTX 2080 Super is only about 4-5% faster than the RTX 2080 Founders Edition, which may not be worth the upgrade if you’ve already invested in the RTX series. However, if you’re upgrading from a GTX card, and you don’t have the funds to shell out for a $1,199 RTX 2080 Ti, the RTX 2080 Super is as close as you’re going to get to robust 4K gaming experience.
The Nvidia GeForce RTX 2070 is less impressive, and while it does absolutely wipe the floor with the GTX 1070, it’s essentially neck in neck with the GTX 1080 – barely hitting a 10% performance increase at 4K in Shadow of the Tomb Raider. At this price point, we expected more, especially after seeing the RTX 2080 and RTX 2080 Ti’s impressive performance.
On the other hand, we’re very impressed with the price-to-performance ratio of the RTX 2070 Super and RTX 2060 Super. This is especially true for the former, which offers a much better performance than the vanilla RTX 2070 at $100 cheaper, and can even breach the RTX 2080’s territory. With boosted specs, the RTX 2060 Super will easily handle 1440p gaming with ray tracing enabled, rivaling the RTX 2070’s performance.
The RTX 2060 is obviously the weakest of the bunch, but you shouldn’t dismiss it outright. The mid-range Nvidia Turing card far outclasses the GTX 1060, but what’s more surprising is that it surpasses the GTX 1070 Ti – for a lower asking price. We were able to get 90 fps in Shadow of the Tomb Raider at 1080p, whereas the 1070 Ti lagged behind at 86 fps. That’s not a huge difference, but the 2060 is $100 cheaper at launch.
In traditional games, there’s no question that Nvidia Turing marks a total upgrade from Pascal. And, over time as drivers mature and users get ready to start overclocking their Turing cards, the difference is only going to grow. That’s not to mention the inclusion of incoming DLSS and ray tracing in games, which should only increase the Nvidia Turing performance gap.
When it comes to ray tracing, there was only one title that supported it at the time of writing: Battlefield V. And, in that title, the Nvidia Turing cards use a hybrid rendering technique – combining both traditional rasterization and ray tracing in order to produce playable frame rates.
Nvidia utilizes “Bounding Volume Hierarchy,” or BVH to track large portions of the scene being rendered for whether or not a ray is being bounced. The RTX cores will then dig deeper into these large rendering zones until it finds the polygon that’s getting hit by the light ray.
This method impacts performance far less than tracking each ray live, but it’s still very demanding. In our own testing, you’ll be stuck at 1080p if you’re looking for smooth gameplay with ray tracing turned on. However, with Nvidia’s latest RTX drivers, it’s claiming to increase performance by up to 50% for ray tracing. We’ll be sure to test this and report back, but we have to wait for the new Battlefield V patch to do it.
As for the Nvidia GeForce GTX 1660 Ti, you can expect a much better performance than the GTX 1060 for less money – up to 56% faster in Shadow of the Tomb Raider at 1080p in our testing. That makes the 1660 Ti a beast when it comes to value.
Kevin Lee is the Hardware and Roundups Editor at IGN Entertainment. Prior to IGN Entertainment, he worked at TechRadar.
Dress right in the hallway. In general, I did not know where it would start. But he surprised me a little. Not forgetting to close the door behind him first, he took a step towards me, dropped to his knees, took my left.
You will also be interested:
- Albertsons cakes
- Maytag 5000 series
- Gresham honda
- Audi q7 wrap
- 6 letter words for kids
- Mechanical engineer salary washington state
- Mesa de otay wait time
- Cross jeep
- Whole food multivitamin by naturelo
- Hetty lange
Good morning to you too. - I grumbled, sinking to the ground next to the girl. - How long have you been sitting. Come see how they will tear me up.