Nvidia GeForce RTX 4090: Release date, price, and specs

• RTX 4090 release date: October 12, 2022
• RTX 4090 price: $1,599
• Built on Nvidia's Ada Lovelace architecture and using TSMC's N4 process
• Nvidia claims the RTX 4090 offers 2–4x performance of the RTX 3090 Ti
• Featuring 16,384 CUDA Cores and 24GB of GDDR6X VRAM

The wait is over. The rumour mill can stop spinning, however briefly, and we can pore over the details Nvidia has released for its new top card. And with Nvidia claiming that the top-end card, the GeForce RTX 4090, is 2-4X faster than its current fastest offering, the GeForce RTX 3090 Ti, it's a generation that is worth paying attention to, even if you can't quite muster the cash to buy one just yet.

We now have the release dates, pricing, and core specifications for Nvidia's new GeForce RTX 40-series GPUs thanks to its GTC 2022 announcement. Built on its Ada Lovelace architecture and produced using TSMC's N4 process (nominally 4nm), there's plenty to get excited about Nvidia's newest silicon. 

Top of the pile is the GeForce RTX 4090, which will be the first card to be released. It's the most expensive offering of those announced so far, but it also packs a mighty punch when it comes, if Nvidia's own benchmarks are to be believed. Here's the lowdown on what's on offer.

Release Date

Only the top of the range GeForce RTX 4090 has a firm release date so far, and that's set for October 12. 

You can sign up to be notified when the cards are available over on Nvidia's site. Resellers are expected not to be taking pre-orders ahead of time either, which is all good by us.

We're expecting review samples ahead of this date because obviously, you wouldn't want to drop serious cash on a card without seeing reviews first. Generally, review embargos lift a few days before the actual launch. 

There's nothing official from Nvidia on this front yet though.

Price

You can expect demand for these cards to be high, particularly at launch, so don't be surprised to see higher prices from resellers. 

The GeForce RTX 4090 has a starting price of $1,599. This makes it $100 more expensive than the GeForce RTX 3090 was at launch, although notably $400 cheaper than the $1,999 RTX 3090 Ti, which is the card Nvidia is comparing the performance of these new models to. 

It's worth noting that both the RTX 3090 and RTX 3090 Ti weren't really aimed at gamers, but more at the semi-professional crowd. This does help make sense of that price point, especially if you're using such cards for rendering and creative work. You know, where time is money.

Performance

We haven't got our hands on the GeForce RTX 4090 yet, so haven't been able to run our own suite of benchmarks on it. In the meantime, Nvidia has released some figures, just to show how its latest architecture compares with its previous best Ampere offering, the RTX 3090 Ti.

In Microsoft Flight Simulator and Warhammer 40,000: Darktide you're looking at a respectable doubling of the frame rate. Nvidia boasts a more than quadrupling of the performance in Cyberpunk 2077 though, albeit using a new ray tracing setting called Overdrive, so we don't have anything we can compare that to just yet. 

The likes of Assassin's Creed Valhalla and Division 2 showing more modest improvements. Still, given the RTX 3090 Ti was no slouch when it came to high-end gaming, this is a good sign for the overall performance. 

Specs

The GeForce RTX 4090 uses Nvidia's new Ada Lovelace architecture, boasts support for the new 3rd generation Ray tracing cores, 4th generation Tensor cores, and DLSS 3. Nvidia has stuck with the PCIe 4.0 interface for this generation and the card support Resizable BAR, Microsoft DirectX 12 Ultimate, and Vulkan RT API.

Nvidia will be producing Founder Editions of the GeForce RTX 4090, although third parties will also be producing their own versions with their own cooling solutions. Such cards can be factory overclocking too, although the difference this makes in game can be slight. 

The GeForce RTX 4090 boasts an incredible 76.3 billion transistors, making it a significantly beefier chip than the existing Ampere GPUs, which rolled in at 28.3 billion.