Nvidia RTX 4070 may not appear this year, but its replacement GPU may surprise you

Nvidia’s RTX 4070 may not be among the first batch of next-gen GPUs released, and if the rumor mill is right, we’ll get two different spins on the RTX 4080 and the flagship RTX 4090.

The latest grapevine chatter on Lovelace comes from multiple sources that VideoCardz (opens in new tab) spoke, combined with information from mainstream hardware leaker Kopite7kimi on Twitter (who recently summarized (opens in new tab) all their latest findings), and Chiphell (opens in new tab) (Chinese forums), all of which confirm VideoCardz’s findings in some places.

There’s some new information about clock speeds and power consumption, but the biggest reveal here is that Nvidia appears to be planning to release two versions of the RTX 4080, one with 16GB of VRAM and one with 12GB – but there will be other big ones. differences with the specification (take all this with large amounts of spice of course, as with any rumor).

The RTX 4080 16GB is the one reportedly built on the AD103 GPU with 9,728 CUDA cores, with the 12GB spin on the 4080 supposedly dropping to use the lesser AD104 GPU, and it will feature 7,680 CUDA cores – quite a difference.

As VideoCardz points out, the RTX 4080 12GB appears to be what was previously identified in the rumor mill as the RTX 4070 (or 4070 Ti) – which explains why it seemed so sturdy, something we’ve noticed in recent times.

VideoCardz informs us that Nvidia has made a “last minute name change” for this card to make it a lesser version of the RTX 4080, and at the moment there is no RTX 4070 – graphics card manufacturers don’t expect this to change (with Lovelace’s launch now so close: it’s expected September 20th at GTC).

Other spec nuggets shared in this new rumor dump is that the RTX 4080 12GB will have a power consumption of 285W (TGP or Total Graphics Power), which can be increased to 366W. The boost clock will reportedly be 2610MHz.

The stronger version of the RTX 4080 with 16 GB of VRAM will reportedly be boosted to 2505 MHz, with a TGP of 340 W (can be pushed to a maximum of 516 W). Specifically, the GDDR6X memory is said to run at 23 Gbps, and this will be the only GPU to have that fast VRAM.

As for the RTX 4090, which boosts to 2520MHz, and the rumor mill points again to that long-held TGP of 450W, but at max it can be pushed to 660W. The 24 GB video RAM will be 21 Gbps modules, the same as the lesser RTX 4080 12 GB.

The tech site tells us that the RTX 4090 should be released in the first half of October, and the RTX 4080 models will apparently debut in the first two weeks of November.


(Image credit: Nvidia)

Analysis: RTX 4070 ETA?

So, what about the RTX 4070 then? Well, we had heard a few weeks ago – again from Kopite7kimi on Twitter – that there were two possible specs for the 4070 graphics card. The mentioned high-end capability is now the RTX 4080 12GB, in theory, so that would leave the other configuration of an AD104 chip with 10GB VRAM and 7,168 CUDA cores the RTX 4070. Perhaps. (Let’s face it, Lovelace’s rumor room is getting more and more confusing).

If this is what is happening, it will be disappointing for a number of reasons. Firstly, with only 10GB of video memory, the RTX 4070 won’t perform well, and secondly, the prospect of it arriving late to the table is also daunting. When does it come up? If not yet in sight, is this a GPU that Nvidia plans to delay until early 2023, as some rumors have suggested?

Before we get too dejected, though, we don’t yet have any indication of pricing for Lovelace outside of the RTX 4090 (we’ve just heard some hints about the flagship’s price tag, and unfortunately the news isn’t what we wanted to hear).

If we don’t get an RTX 4070 initially, but we get an RTX 4080 12GB which is relatively well priced for its performance – considering it’s a decent distance from the 4080 16GB in terms of a lesser GPU and significantly fewer CUDA Cores – then it makes it might not matter. That is, it’s all about the price/performance ratio of these GPUs, whatever they may be called.

That Nvidia has supposedly decided to elevate the name to an RTX 4080 from the 4070, however, leaves a disturbing feeling about the pricing in our bellies. Especially given the RTX 4090’s aforementioned price leak. Time will tell, as always, and with less than a week to go before Lovelace’s launch, we won’t have long to wait to discover more about the next generation of graphics cards from Team Green.

However, all of this leaves us wondering if the tactic might be to charge a premium for the new products – pretty much as expected for high-end GPUs fresh off the assembly line – and the prices of that excess RTX 3000 stock. lower it further as the alternative option for those who don’t want to splash at a lot.