Intel is preparing its first processors for PC laptops to integrate the RAM directly on the chip for lighter, but less scalable PCs, in the manner of Macbook.
Like other manufacturers, Intel is preparing to integrate more and more elements into its processors. GPUs and other NPUs are already on the Meteor Lake menu.
Lunar Lake will, however, go a little further and integrate the system’s RAM within the chip. A major technological development that will bring Intel a little closer to Apple.
Following the path taken by Apple
While Apple is obviously not the only one, it remains one of the best-known processor designers to integrate RAM into its chips. Thus, from 2020, the M1 processor had LPDDR4X SDRAM.
At Intel, things came very gradually and if the idea may have been in the minds of its engineers since Meteor Lake, it is indeed the Lunar Lake generation – the release of which is planned before the end of the year 2024 – which will usher in this technological innovation. The information relayed by VideoCardz was revealed by InstLatX64.
The latter discovered the existence of a new reference in the technical data around Lunar Lake processors, the Core Ultra 5 238V which therefore joins the already noticed Core Ultra 5 234V.
Characteristics of the memory integrated into Lunar Lake CPUs © Intel
Up to 32 GB of memory… non-scalable
The two chips come from this new – the second – generation of Intel chips designed for very low consumption (8 to 30 watts) and, therefore, integrating LPDDR5X SDRAM memory.
In this case, we are talking about LPDDR5X-8533 for which two configurations are possible depending on whether Intel integrates 16 GB or 32 GB. Logically, the Core Ultra 5 234V will be equipped with 16 GB when the Core Ultra 5 238V will have 32 GB of RAM. This will also be the essential difference between these two chips since the number of CPU (4C+4c) and GPU (7 Xe2-Cores) cores is identical. That said, perhaps the frequencies will differ.
The integration of RAM into the processor makes it possible to produce machines that are easier for manufacturers to manufacture, more compact and lighter. On the other hand, this makes it completely impossible to increase this amount of memory.
Source : InstLatX64, VideoCardz
5