Nvidia RTX 4000 graphics cards are officially here to dethrone the current-gen GeForce RTX 3000 GPUs, we’ll need to wait a little longer for the entire cavalry to arrive. To help you navigate the convoluted waters of announcements, leaks, and rumours, we’ve compiled everything we know so far about the next-generation pixel pushers.
So far, all of the next-gen best graphics card contenders cater to desktops, but we’ll apparently get an RTX 4090 gaming laptop GPU further down the line. Naturally, cheaper mobile options like the RTX 4050 are also on the cards, and the entire lineup will benefit from Nvidia DLSS and Reflex.
As much as RTX 4000 GPUs undoubtedly offer a high level of performance, Lovelace faces stiff competition from the likes of AMD RDNA 3. Thankfully, the RTX 4090 and RTX 4080 are already a thing, and we shouldn’t have to wait much longer to find out how the RTX 4070 and RTX 4060 stack up against their rivals.
Here’s everything we know about the Nvidia RTX 4000 GPUs:
Nvidia RTX 4000 release date
The RTX 4090 released on October 12, while its RTX 4080 sibling rocked up to the show on November 15, 2022. We’re still waiting for the entire Lovelace lineup to arrive, but we’ll likely hear more about the Nvidia RTX 4070 Ti in early 2023.
Various RTX 4000 model names are now listed as an EEC trademark, which doesn’t guarantee an ETA, but it does bolster claims that they’ll show up in a few months’ time. Just don’t expect the RTX 4090 Ti to pop up anytime soon, as it seems Nvidia needs to stop the GPU from melting before sending it to market.
Nvidia RTX 4000 price
The Nvidia RTX 4090 carries a hefty $1,599 USD price tag, which is only $100 more than the RTX 3090 MSRP two years prior. The RTX 4080 is priced at $1,199 USD / £1,269 GBP, which is nearly double the RTX 3080’s $699 starting price and even more than the 12GB RTX 3080 Ti, which originally cost $1,199 USD. Of course, these prices skyrocketed due to a global chip shortage, but it’s surprising to still see such high costs given that particular issue is longer over.
The RTX 4060 and RTX 4070 still aren’t official, so we can only speculate how much they’ll cost using the flagships as a starting point. We expect the RTX 4070 could arrive at around $549 – $699 USD, while the RTX 4060 might sit near the $349 – $500 USD mark.
Comparatively, the AMD Radeon RX 7900 XTX costs $999, and the Radeon RX XT sits at $899. Neither can keep up with the monstrous abilities of the RTX 4090, but the red team’s flagship trades blows with the RTX 4080.
Nvidia RTX 4000 specs
The Nvidia RTX 4000 specs were originally a bit confusing because of two different RTX 4080s. Fortunately, team green thought better of itself and has since narrowed it down to one.
On a broader scale, all RTX 4000 GPUs run on the new Ada Lovelace architecture, which improves the RTX processors and neural rendering engines. They also support the new and improved DLSS 3.0 upscaling, which doesn’t work on older GPUs. Eventually, this should help revamp the best gaming laptop scene, if mobile RTX 4000 graphics cards are announced at CES 2023.
The GeForce RTX 4080 graphics card has 16GB GDDR6X VRAM, which is a good starting point since games like Far Cry 6 have some trouble with anything less than 12GB at higher resolutions and settings (such as ray tracing). That said, AMD Radeon GPUs have been clocking 16GB since its RTX 3000 rivals hit the market. There was going to be an RTX 4080 12GB version, but this is no longer a thing.
Nvidia says that the RTX 4080 boasts between two and four times the speed of an RTX 3080 Ti on RacerX. It’s worth mentioning that Nvidia CEO Jensen Huang doesn’t position these as replacements for the RTX 3000 series but as a complementary enthusiast range that supplements it.
The RTX 4090 clocks in with a whopping 24GB of GDDR6X VRAM, leading the charge for Ada Lovelace. While it’s expected to guzzle on your electric, Nvidia recommends an 850W power supply, so we haven’t yet reached the clutches of four digits.
According to reliable leaker Kopite7kimi, an Nvidia RTX 4090 Ti that uses a “full fat” version of the AD102 GPU could be on the cards. The insider does specify that it’s unclear whether it’ll become an actual product, but the souped up graphics card could come equipped with 46GB GDDR6 VRAM while requiring two 16-pin power connectors.
Full fat AD102 GPU performance claims have also surfaced online, and insiders say the Nvidia RTX 4090 Ti can run Control at 4K 160fps with DLSS and ray tracing enabled.
CUDA Cores | L2 Cache | |
AD102 | 18,432 | 96MB |
---|---|---|
AD103 | 10,752 | 64MB |
AD104 | 7,680 | 48MB |
AD106 | 4,608 | 32MB |
AD107 | 3,072 | 32MB |
Further leaks point towards enhanced AD104 specs, as the RTX 4070 apparently packs RTX 3090 Ti levels of performance. It also wields faster clock speeds than the RTX 4090, which could make it an excellent mid-range option.
Nvidia RTX 4000 benchmarks
RTX 4000 benchmarks have arrived, and our Nvidia GeForce RTX 4090 review puts the flagship Lovelace leviathan to the test. The next-gen can boost fps beyond the limits of the RTX 3090, even with ray tracing on and ultra presets enabled.
Nvidia DLSS 3.0 plays a huge role in uplifting the RTX 4090 GPU’s capabilities, and frame generation is a game changer. While it’ll take a while for support to roll out across the board, the tool pushes Cyberpunk 4K ray tracing frame rates into triple-digit territory.
Of course, the RTX 4090 won’t be the only GPU armed with DLSS 3.0, as the entire RTX 4000 range should boast compatibility with the tech. If flagship upscaling performance is anything to go by, even entry-level cards like the RTX 4060 may pack enough of a punch to match previous high-end contenders.
Watch this space for our RTX 4080 review, as the next high-end Lovelace graphics card is scheduled to arrive in November.