AMD's upcoming Zen 6-based Medusa Point mobile processors won't be sharing the same FP8 socket as used by Strix Point.
NBD Shipping Manifest Reveals Medusa Point's Socket Details: Will Utilize FP10 Socket, Measuring 25mm x 42.5mm
AMD's Medusa Point mobile processors, unlike previously thought, won't be using the predecessor's FP8 socket that is currently being used by AMD's Zen 5-based Strix Point CPUs. While the detailed specifications and features of Medusa Point are yet to be known, we have gradually grabbed more and more information in recent days, the socket being the most recent one.
As spotted by the user @Olrak29_, the Medusa Point and its socket have been leaked in the NBD shipping logs, detailing what it will bring to the table compared to what we currently have on the Strix Point. Contrary to what was previously thought, AMD's Medusa Point won't be using the same FP8 socket as Strix Point but will use the slightly bigger FP10 socket. The size is going to be 25mm x 42.5mm, which is roughly 6% bigger than FP8.
Such details might not be very useful for the masses, but it reveals that the Medusa Point processors might be slightly bigger than the Strix Point CPUs. Strix Point CPUs feature a die size of 232.5mm2, bringing 12 cores based on Zen 5 and Zen 5c architectures. The Medusa Point is expected to utilize a single 12-core chiplet based on Zen 6 architecture.
The die design is going to be quite different from that of Strix Point CPUs that do it with a combination of 4+8 cores on a monolithic die, featuring all other components on a single silicon die. On the other hand, the Medusa Point is expected to bring a chiplet-based design, featuring a dedicated CCD for hosting 12x Zen 6 cores and a separate I/O die. The next major difference is the process node with Medusa Point being manufactured on TSMC's 3nm process node compared to the 4nm used by Strix Point.
As far as the integrated graphics go, there have been very few reports detailing what Medusa Point will use, but we recently confirmed that the iGPU on Medusa Point won't use RDNA 4 architecture. It will carry on with the RDNA 3.5 while the RDNA 4 will be dedicated only to the discrete GPUs. Future AMD mobile generations are expected to utilize the more advanced UDNA architecture, but Medusa Point isn't one of them.
The release date is also unclear, but it's expected to be released next year. We await more such leaks and official confirmation from AMD on Medusa Point and will keep you updated with the latest information.
