Getting your hands on a powerful graphics card is still very difficult. That isn’t stopping NVIDIA, the company behind some of the most desirable GPUs and processors, from rolling out even more of them. As reported by DigiTimes (paywall), the electronics company and its partners (namely Taiwanese factories that manufacture the necessary components) are gearing up to announce the release of a next-gen GPU for 2022.
The GPU in question — a GeForce RTX 40 series graphics card based on the Ada Lovelace GPU architecture — is rumored to have almost twice the power of the current line of RTX 3080 chips. To put into context, the RTX 3080 series boasts 34 terraflops (TFLOPs) of power, while the newer models are said to feature 64 TFLOPs.
According to DigiTimes, its report is based on information coming directly from NVIDIA factory partners in Taiwan. And while you have every right to be excited about all the things you’re going to do with these beastly chips, you should also get ready to spend a serious amount of time, patience, and funds trying to get one — much like with the RTX 3080. Because, at this point, it’s easier to access a cloud server that is powered by RTX 3080s than buying an individual one.
Console comparison— As broken down by Gizmodo, NVIDIA’s rumored next-gen GPU stacks up comfortably against the power of the current slate of next-gen consoles. For example, the Xbox Series X is the strongest one on the market and features 12 TFLOPs of power. The Playstation isn’t too far behind, maxing out at 10.3 TFLOPs. Both of these figures pale in comparison to the supposed power of the RTX 40.
Next year is shaping up to be a busy one for NVIDIA: On top of the RTX 40 buzz, the company is planning to drop a new line of souped up RTX 30s through variations on existing models, like the 12GB RTX 3080, 16GB RTX 3070 TI, and 12GB RTX 3060. As such, it makes sense that even if an RTX 40 announcement is made, it might not coincide with a release until next holiday season when supply chain issues may be less prominent and the revamped line of RTX 30s has had time to breathe.
Tempered expectations — The ADA Lovelace architecture will employ a smaller node size of 5mm (compared with 8mm) and be capable of up to 71 percent faster performance when compared to chips constructed out of Ampere architecture, which powers the RTX 30s.
Still, even with the news it’s hard to imagine a world where acquiring a next-gen GPU will be anything but a migraine. All we can do is wait.