You’ll no doubt remember the announcement of the collaboration between Intel and Nvidia earlier this year, which took quite a few people by surprise. During the official announcements, information on future products linked to this collaboration remained very nebulous. An X86 SoC with graphics developed with Nvidia was clearly mentioned, however. We now know a little more, but we’ll have to be patient, according to the leaks that RedTechGaming claims to have in its possession.
According to the leaks, this is a graphics solution based on future NVIDIA architectures (such as possibly Rubin) engraved at TSMC, combined with LPDDR6 memory, with a configuration of up to 16 GB LPDDR6. The ambition would be to offer an integrated solution featuring a powerful, efficient CPU combined with a graphics-level part offering much higher performance than current integrated solutions, capable of greatly reducing – or even eliminating – the need for a dedicated graphics card on a greater number of notebook PCs. Obviously, AMD SoCs are in the crosshair of this project.
Serpent Lake not before 2029: too late?
For the CPU part, Intel would use Titan Lake architecture (with Griffin Cove P cores and enhanced Golden Eagle E cores). The SoC tile, which houses the LP E-cores, could also be based on Intel’s Titan Lake technologies. Speculation is still rife, but rumors point to support forLPDDR6 and LPDDR5X memory .
Obviously, if we look at the current memory situation, a chip integrating 16GB of LPDDR6, that’s an immediate cause for concern.
In the video, to be taken with great caution of course, the author also talks about the succession to Nova Lake (which won’t arrive until mid-2026). Razor Lake and Titan Lake are also discussed, although once again we’re really in the realm of the crystal ball.
