NVIDIA today introduced the GeForce RTX 40 series GPUs, which offer revolutionary performance for gamers and creators. First and foremost, the new flagship RTX 4090 offers up to four times the performance of its predecessor.

As the world’s first GPU based on the new NVIDIA Ada Lovelace architecture, the RTX 40 series offers a significant generational shift in terms of performance and efficiency. The series represents a new era of real-time ray tracing and neural rendering that uses AI to generate pixels.

“The era of RTX ray tracing and neural rendering is in full swing, and our new Ada Lovelace architecture takes it to the next level,” said Jensen Huang, NVIDIA Founder and CEO, on the GeForce Beyond: Special Broadcast at GTC.

“Ada offers a quantum leap for gamers and paves the way for fully simulated worlds for developers. With up to four times the performance of the previous generation, Ada sets a new standard for the industry,” he says.

DLSS 3 generates whole frames for faster gameplay

Huang also announced NVIDIA DLSS 3 – the next revolution in the company’s proprietary deep learning super sampling neural-graphics technology for gaming and creative applications. The AI-powered technology can generate entire frames to massively speed up gameplay. It can overcome CPU-related performance limitations in games by allowing the GPU to generate entire frames independently.

The technology will be integrated into the world’s most popular game engines such as Unity and Unreal Engine. Many of the world’s leading game developers support them. More than 35 games and apps will be released soon.

In addition, the RTX 40 series GPUs offer a number of new technological innovations, including:

  • Streaming multiprocessors with shader performance of up to 83 teraflops – more than double the previous generation
  • Third-generation RT cores with up to 191 teraflops for effective ray tracing – 2.8 times more than the previous generation
  • 4th generation Tensor cores with up to 1.32 tensor petaflops – 5x more than the previous generation with FP8 acceleration
  • Shader Execution Reordering (SER) improves execution efficiency by rescheduling shading workloads on the fly. This makes better use of the GPU’s resources. As significant an innovation as out-of-order execution for CPUs, SER improves ray tracing performance by 3x and frame rates in games by up to 25%
  • Ada Optical Flow Accelerator with twice the performance enables DLSS 3 to predict motion in a scene, allowing the neural network to increase frame rate while maintaining image quality
  • Architectural improvements tightly coupled with custom TSMC-4N process technology result in up to a 2x improvement in power efficiency

Two NVIDIA encoders (NVENC) reduce export times by up to 50% and offer AV1 support. NVENC AV1 encoding is used by OBS, Blackmagic Design DaVinci Resolve, Discord and many others.

New ray tracing technology for even more immersive games

For decades, rendering ray-traced scenes with physically correct lighting in real time has been considered the holy grail of graphics. At the same time, the geometric complexity of environments and objects has continued to increase as 3D games and graphics strive to represent the real world as accurately as possible.

Physically correct graphics require enormous computing power. Modern ray tracing games like Cyberpunk 2077 perform over 600 ray tracing calculations on each pixel just to determine the lighting – a 16x increase over the first ray tracing games launched four years ago.

The new third generation RT cores have been enhanced to enable twice as fast ray triangle intersection tests and include two important new hardware units. An “Opacity Micromap Engine” speeds up ray tracing of alpha test geometries by 2x, and a “Micro-Mesh Engine” creates micro-meshes on the fly to generate additional geometry. The micro-mesh engine offers the benefits of increased geometric complexity without the traditional performance and memory costs of complex geometries.

Creativity redefined with RTX Remix, new AV1 encoders

The RTX 40 series GPUs and DLSS 3 offer new possibilities for NVIDIA studio developers. 3D artists can render fully ray-traced environments with accurate physics and realistic materials, and view the changes in real time and without proxies.

Video editing and live streaming also benefit from improved GPU performance and new dual 8th generation AV1 encoders. There are three updates to the NVIDIA Broadcast Software Development Kit that are now available for partners. These include face expression estimation, eye contact, and virtual background quality improvements.

NVIDIA Omniverse, part of the NVIDIA Studio software suite, will soon be expanded to include NVIDIA RTX Remix, a modding platform for creating stunning RTX remasters of classic games. With RTX Remix, modders can easily capture game assets, automatically enhance materials with powerful AI tools, and quickly enable RTX with ray tracing and DLSS.

portal is RTX ON!

Remix RTX was used by NVIDIA Lightspeed Studios to redesign Valve’s iconic video game Portal, considered one of the best video games of all time. Advanced graphics features like full ray tracing and DLSS 3 give the game a stunning new look and feel. Portal w/ RTX will be released as free, official downloadable content for the classic platformer with RTX graphics in November, just in time for Portal’s 15th anniversary.

The GeForce RTX 4090 and 4080: The new ultimate GPUs

The RTX 4090 is the world’s fastest gaming GPU with amazing performance, acoustics and temperature characteristics. In fully ray-traced games, the RTX 4090 is up to 4x faster with DLSS 3 than the last-gen RTX 3090 Ti with DLSS 2. It’s also up to 2x faster in current games while maintaining 450W power consumption. It features 76 billion transistors, 16,384 CUDA cores and 24GB of high-speed Micron GDDR6X memory, delivering consistently over 100 frames per second for 4K resolution games. The RTX 4090 will be available starting October 12th, starting at $1,599.

The company also announced the RTX 4080, which will be available in two configurations. The RTX 4080 16GB has 9,728 CUDA cores and 16GB of high-speed Micron GDDR6X memory and with DLSS 3 it is twice as fast as the GeForce RTX 3080 Ti in current games and more powerful than the GeForce RTX 3090 Ti with lower power consumption. The RTX 4080 12GB features 7,680 CUDA cores and 12GB of Micron GDDR6X memory, and with DLSS 3, it’s faster than the RTX 3090 Ti, the flagship GPU from the previous generation.

Both RTX 4080 configurations will be available in November with prices starting at $1,199 and $899, respectively.

Where to buy them

The GeForce RTX 4090 and 4080 GPUs will be available in standard clock custom boards and factory overclocked models from top add-in partners including ASUS, Colorful, Gainward, Galaxy, GIGABYTE, Innovision 3D, MSI, Palit, PNY, and Zotac .

The RTX 4090 and RTX 4080 (16GB) are also produced directly by NVIDIA in limited Founders Editions for fans of NVIDIA’s own design.

The GeForce RTX 40 series GPUs will be found in gaming systems from Acer, Alienware, ASUS, Dell, HP, Lenovo and MSI, leading system manufacturers worldwide, and many other manufacturers.