The rumour was first hinted by prominent and reliable leakster, kopite7kimi (@kopite7kimi), who kind of confirmed that small detail with a reply to another source’s tweet. Apparently, NVIDIA’s launch strategy for this round will be for the company to let the very top-tier of its Ada Lovelace architecture pave the way for its next generation of graphics cards. By that, they mean that NVIDIA will be launching the RTX 4090 first, followed by the RTX 4080, and then the RTX 4070. As to why the launch sequence is mildly important: back when NVIDIA launched its Ampere architecture cards, it announced the RTX 3090, RTX 3080, and RTX 3070 all in the same day. However, the RTX 3080 was the first card to reach our hands, while the RTX 3090 and RTX 3070 only became available to us between a period of two weeks and two months, respectively.
— kopite7kimi (@kopite7kimi) May 29, 2022 Unfortunately, kopite7kime’s disclosure seems to end there, but not before 3DCenter.org (3DCenter_org) confirmed yet another small detail: that the full-fat AD102 GPU will be made available much, much later. It all sounds par for the course, considering that NVIDIA may simply be waiting to see just how formidable AMD’s own Navi 31 architecture would be. To quickly recap on other specs, NVIDIA’s GeForce RTX 40 series is expected to shift from the use of PCIe Gen4 to PCIe Gen5. Not only that but its cards are also rumoured to be capable of pulling as much as 600W off the wall, all through a single and unique 16-pin microFit MOLEX PCIe Gen5 connector. In any case, these are all just rumours, so do treat them with the discretion required with such matters. (Source: Twitter, Videocardz)
