Nvidia’s GeForce RTX 30 Series offered a significant jump in performance compared to the RTX 20 Series, and rumours suggest we could be seeing a similar jump from the next-gen RTX 40 Series currently in development behind closed doors at Nvidia HQ.

The latest leaks point towards a 5nm process node running Nvidia’s Ada Lovelace GPU architecture that could offer a 2x jump in performance, and whispers suggest release could be closer than expected. Here’s all there is to know about the Nvidia GeForce RTX 40 Series right now.

When will the Nvidia RTX 40 Series be released?

Nvidia’s next-gen RTX 40 series of graphics cards are due for release sometime in late 2022 – and given the two-year gap between the RTX 20 series and RTX 30 series, that shouldn’t come as much of a surprise.

According to Twitter leaker @Greymon55, the graphics cards will launch "no earlier than end of 2022" which aligns with previous Nvidia graphics card releases, which tend to appear towards the end of the year, although the leaker has also suggested that this could potentially slip into Q1 2023, which runs from January – March.

The latter makes a lot of sense, especially with Nvidia confirming a 2023 release of its Grace server chip based on the same rumoured 5nm process as the RTX 40 Series.

A Q1 2023 release could be a good thing, hopefully giving Nvidia (and everybody else) enough time to get over the current component shortage causing manufacturing chaos throughout the tech industry.

However, a DigiTimes report suggests a slightly earlier release could be on the cards. Citing industry sources (we've heard that ol' chestnut before!) the publication suggests that the 5nm RTX 40 Series will be released in 2022, though didn't provide a specific window. However, given DigiTimes hit-and-miss track record when it comes to leaks, we'd take this with a big pinch of salt.  

It is worth noting that reliable leaker @Kopite7kimi has suggested that the AD102 GPU - the GPU expected to power the likes of the RTX 4080 and 4090 - entered the testing phase in April 2022. If true, it's a good sign that development of the next-gen range is coming along nicely, and barring any unforeseen testing issues, the design of the GPU is likely finalised.

The leaker has followed up his claim, teasing in May 2022 that a release could be on the cards as soon as early Q3, which begins in July. Given Nvidia isn't expected to announce the next-gen graphics cards at Computex in late May, a release in August - possibly around Gamescom - could make more sense. 

How much will the Nvidia RTX 40 Series cost?

We’re still some time away from the release of the Nvidia RTX 40 Series, so it’s hard to say how much the next-gen range might cost right now – although we can look at the current RTX 30 Series as a guide.

As a reminder, here’s what the current RTX 30 Series line-up looks like:

Starting at just £299/$329, Nvidia’s RTX 3060 comes in cheaper than competing GPUs from AMD’s Radeon RX 6000 Series, and it’s hoped that Nvidia will keep that trend going with the next-gen 40 Series – after all, there were significant jumps in performance in the 30 Series compared to its predecessor and it was still competitively priced.

The issue, as most of you reading this article will be painfully aware of, is the lack of stock, driving GPU prices through the roof with resellers marking up prices to never seen before levels – it’s not uncommon to see an RTX 3070, a £469/$499 card, go for double that amount. The worst part? People are actually paying those prices.

Nvidia could see that gamers are willing to pay much more than the MSRP for its graphics cards, which could incentivise the company to increase the price across the range for the RTX 40 Series – something rumoured to be happening with the competing AMD Radeon RX 7000 Series.

Let’s hope that doesn’t happen though; I think we all need a break from pricey GPUs after the current shortage ends.

The good news is that Nvidia is bolstering stock of the RTX 30 series, with Nvidia CFO Colette Kress confirming at a March 2022 investors event that "we will be in a good supply situation in the second half [of 2022]" which should translate to better pricing.

Kress also teased the idea of continuing to sell the RTX 30 series alongside the 40 series to provide more choice - and crucially, supply - for consumers.

"Even during this period of COVID and supply constraints, it’s been interesting because it’s given us the opportunity for gaming to continue to sell both the current generation (RTX 30 series) as well as the Turing generation (the RTX 20 series)," Kress explained. "So we’ve been doing that to provide more and more supply to our gamers in that. And we may see something like that continue in the future."

What to expect from the Nvidia RTX 40 Series

5nm Ada Lovelace architecture

Twitter leaker @Greymon55 suggests that Nvidia’s GeForce RTX 40 Series will be based on the Ada Lovelace GPU architecture – a figure associated with the creation are what are considered to be the first computer programs back in the 1800s for those unaware.

It’s also suggested that the new range will be based on TSMC’s 5nm process node, down from 8nm on the current range. What isn’t known right now is whether it’s based on the standard N5 or N5P node.

The 5nm process has been outed a number of times from varying reports - most recently from DigiTimes - suggesting that there could be weight to the claim. 

Performance improvements

Despite the huge gains in performance the RTX 30 Series offered over the 20 Series, the RTX 40 Series is expected to offer a similar boost in performance when compared to the current top-end RTX 3090 – a graphics card with an MSRP of £1399/$1499.

The rumour comes from Twitter leaker @TtLexington, who suggests that the Nvidia RTX 40 Series could offer the same general performance jump over the RTX 30 Series, offering a rough 2x jump in overall performance, courtesy of the Ada Lovelace GPU architecture and 5nm chipset at the heart of the range.

That backs up an earlier rumour that suggests the Lovelace AD102 GPU computes at an impressive 66.4 TFLOPs, compared to the 38.7 TFLOPs of compute power on offer from the current Ampere GA102 chip. It also suggests a boosted 144 streaming multiprocessors and a maximum 18,432 CUDA cores – almost double that of the RTX 3090.

It’s also rumoured that the new RTX 40 Series could feature clock speeds between 2.2 and 2.5GHz (boosted), offering a decent jump from the current 1.7-1.9GHz clock speeds on offer from the current Ampere architecture.

t’s still not enough to beat the 2.5GHz on offer from AMD’s competing RDNA 2 architecture available right now, however, let alone what AMD could be cooking up with its upcoming RDNA 3-based range.

Related content