There's no longer any question of launching them early next year, but rather around the summer of 2026, or even in September.
For several weeks now, several indicators have been confirming that the supply of memory chips (DRAM and NAND) is very inadequate, while demand is exploding due to the growth of the artificial intelligence market. It's no secret that AI is very greedy for storage space, whether we're talking about memory or hard disks/SSDs. The problem is that this greed has consequences for all other markets.
Indeed, with demand far outstripping supply, memory chip manufacturers naturally turn to those customers willing to pay the most for this precious commodity. As a result, all AI-related projects are taking priority, and other sectors are paying the price. In recent weeks, the price of RAM modules has risen sharply, and SSDs may well follow suit.
On X.com, MEGAsizeGPU specifies the estimated launch window for the GeForce RTX 50 SUPER series © MEGAsizeGPU
For NVIDIA, this means higher graphics card manufacturing costs for partners such as ASUS, Gigabyte and MSI, to name but three of the most important. The problem is that several rumours indicate that NVIDIA is planning to launch new graphics cards - the GeForce RTX 50 SUPER series - at the beginning of next year, in order to keep up with its usual launch/update rhythm. Indeed, the GeForce RTX 50 series was launched last January.
These new "SUPER" graphics cards seemed set to focus on increasing on-board video memory, which isn't exactly a good thing when we've just explained how tight the memory market is at the moment... and for several months to come. NVIDIA has therefore decided to postpone the launch of its RTX 50 SUPER series cards, which are no longer expected before the third quarter of 2026: perhaps time to stock up on memory chips or for the market to calm down?
