As part of the keynote presentation for NVIDIA’s GTC conference, NVIDIA founder and CEO Jensen Huang swapped the kitchen backdrop seen during the RTX 30 Series announcement for something a little more cosmic. After showcasing an impressive interactive tech demo built using the latest in RTX technologies we got confirmation that the GeForce RTX 4090 and GeForce RTX 4080 graphics cards will launch in October and November of this year. And yeah, they’re beasts.

And with the prices of the GeForce RTX 30 Series coming down, the new 40 Series cards are targeting enthusiasts. In other words, they’ll be pricey ala the RTX 3080 and RTX 3090 launches. With the RTX 4090 starting at $1,599 USD and the RTX 4080 (16GB) USD at $1,199 and RTX 4080 (12GB) at $899 USD - two models for the RTX 4080 that feature different specs.

Of course, the cost here comes from some truly impressive hardware, with NVIDIA boasting 2X to 4X the performance for the GeForce RTX 4090 compared to the current beast, the GeForce RTX 3090 Ti. Thanks to the advancements made with 3rd generation RTX hardware, the new Ada Lovelace architecture built on TSMC 4N, and the impressive DLSS 3, sounds pretty incredible.

NVIDIA made the announcement by showcasing the fully ray-traced interactive demo (that will launch this November) Racer RTX. As it’s built on simulating real-world objects, this was apparently put together in the span of only a few months.

Racer RTX showcases the latest NVIDIA technologies including real-time ray tracing, DLSS 3, and PhysX. Available as a playable tech demo for GeForce RTX 40 Series GPUs this November, Racer RTX is an interactive physics-accurate simulation featuring the most realistically rendered RC cars ever.

And here’s all the numbers you need to know relating to the new GeForce RTX 40 Series.

GeForce RTX 4090

---

NVIDIA Architecture: Ada Lovelace
NVIDIA CUDA Cores: 16384
Boost Clock (GHz): 2.52
Base Clock (GHz): 2.23
Memory: 24 GB GDDR6X
Memory Interface: 384-Bit
Ray Tracing Cores: 3rd Gen
Tensor Cores: 4th Gen
Length/Width/Height: 304x137x61mm (3-Slot)
Power: 450 W
Price: Starting from $1,599 USD and $2,959 AUD

GeForce RTX 4080

---

GeForce RTX 4080 (16GB)

NVIDIA Architecture: Ada Lovelace
NVIDIA CUDA Cores: 9728
Boost Clock (GHz): 2.51
Base Clock (GHz): 2.21
Memory: 16 GB GDDR6X
Memory Interface: 256-Bit
Ray Tracing Cores: 3rd Gen
Tensor Cores: 4th Gen
Length/Width/Slot: 304x137x61mm (3-Slot)
Power: 320 W
Price: Starting from $1,199 USD and $2,219 AUD

GeForce RTX 4080 (12GB)

NVIDIA Architecture: Ada Lovelace
NVIDIA CUDA Cores: 7680
Boost Clock (GHz): 2.61
Base Clock (GHz): 2.31
Memory: 12 GB GDDR6X
Memory Interface: 192-bit
Ray Tracing Cores: 3rd Gen
Tensor Cores: 4th Gen
Length/Width/Slot: TBC
Power: 285 W
Price: Starting from $899 USD and $1,659 AUD

With massive performance gains compared to the RTX 30 Series, it seems like the impressive sounding DLSS 3 will be exclusive to the GeForce RTX 40 Series. This doesn’t sound great for those with 30 Series GPUs, but it sounds like DLSS 3 is only possible thanks to the new Optical Flow Accelerator - which uses AI to generate frames and even take the burden off of the CPU. NVIDIA showcased with this look at the CPU-intensive Microsoft Flight Simulator, where DLSS 3 more than doubles in-game performance.


It seems the most significant gains that the Ada-powered GPUs bring versus the Ampere-powered RTX 30 Series will come with games with ray-tracing - where NVIDIA was on-hand to showcase a 4X increase for Cyberpunk 2077 using a brand-new ray-tracing mode that will take the already hardware intensive game and add even more ray-traced detail to environments.

Cyberpunk 2077’s neon-illuminated environments are key to its aesthetic, and with the new Ray Tracing: Overdrive Mode their level of detail is taken to the next level:

NVIDIA RTX Direct Illumination (RTXDI) gives each neon sign, street lamp, car headlight, LED billboard, and TV accurate ray-traced lighting and shadows, bathing objects, walls, passing cars, and pedestrians in accurate colored lighting.
Ray-traced indirect lighting and reflections now bounce multiple times, compared to the previous solution’s single bounce. The result is even more accurate, realistic and immersive global illumination, reflections, and self-reflections.
Ray-traced reflections are now rendered at full resolution, further improving their quality. And Improved, more physically-based lighting removes the need for any other occlusion techniques.

This is achieved through the new RT Cores, Tensor Cores, and advances that power DLSS 3. But also something that NVIDIA is calling Shader Execution Reordering (SER) which presents a massive boost to GPU workflow. Here’s the skinny.

GPU architecture is highly parallelized and at its most efficient when executing similar workloads simultaneously. However, advanced ray tracing requires computing the impact of millions of rays striking numerous different material types throughout a scene, creating a sequence of divergent, inefficient workloads for shaders (shaders calculate the appropriate levels of light, darkness, and color during the rendering of a 3D scene, and are used in every modern game).

Our new Shader Execution Reordering (SER) technology dynamically reorganizes these previously-inefficient workloads into considerably more efficient ones, improving shader performance by up to 2X, and in-game frame rates by up to 25%.

Yeah, pretty impressive that this new tech can boost frame rates on its own by 25%.

Finally, in addition to the Founders Edition reveals NVIDIA has noted that several partner cards will be made available at launch too. For more information head here.