Recent foreign media reports have exposed the core configuration of AMD’s next-generation Zen 6-based mobile processor series, code-named Medusa Point. The leaked information indicates that the flagship Ryzen 9 models in this series will feature up to 22 cores, promising a significant boost in multi-threaded performance. However, there appears to be a compromise in the GPU configuration, with the integrated graphics (iGPU) adopting the RDNA 3.5+ architecture and the number of Compute Units (CUs) reduced to 8, a step back compared to previous generations.
Medusa Point Series: A Significant Mobile APU Update #
The Medusa Point series represents a major generational upgrade for AMD’s mobile APUs, positioned as the main mobile product line following the Zen 5-based Strix Point and Strix Halo. The Ryzen 5 and Ryzen 7 models will feature a 10-core design, comprising 4 Classic Cores, 4 Dense Cores, and 2 Low-Power Cores, paired with an iGPU utilizing 8 RDNA 3.5+ Compute Units. The Ryzen 9 model is more aggressive, achieving a total of 22 cores by adding an additional 12-core Chiplet Complex Die (CCD) on top of the 10-core monolithic die, while the iGPU still maintains 8 Compute Units. This Multi-Chip Module (MCM) design marks AMD’s first attempt to bring desktop-class high core count chips to the laptop platform, aiming to enhance multi-threaded task processing capabilities.
Zen 6 Architecture: The Key to Performance Leap #
The Zen 6 architecture is crucial for Medusa Point’s performance leap. Unlike previous Zen architectures where a single chiplet complex supported a maximum of 8 cores, Zen 6 increases the core count per chiplet complex to 12. It also introduces a hybrid architecture with three types of cores: Classic, Dense, and Low-Power. This design may partially draw inspiration from Intel’s hybrid core strategy, aiming to balance performance and energy efficiency. While many have expressed skepticism about Intel’s big.LITTLE design, it’s proving to be a future hardware trend, and with perfect software scheduling, it’s definitely a preferred option for home PCs.
Returning to the Zen 6-based Medusa, among the three core types, Classic Cores are responsible for high-performance computing tasks like game rendering and video encoding; Dense Cores target multi-threaded light tasks, providing efficient parallel processing; and Low-Power Cores focus on background tasks and low-load scenarios to effectively extend battery life. The 22-core Ryzen 9 model is particularly suitable for mobile workstations requiring extremely high multi-threaded performance, such as 3D modeling, virtualization, or scientific computing.
Packaging and Modular Design #
In terms of packaging technology, Medusa Point adopts the FP10 package, which is slightly larger than Strix Point’s FP8 package, measuring approximately 25mm x 42.5mm, an area increase of about 6%. The larger package provides more space for thermal design and internal component layout, helping maintain high-performance output in thin-and-light laptops. Some Medusa Point models use a monolithic chip design, integrating 10 cores and the iGPU, while the Ryzen 9 model connects an additional 12-core chiplet complex to the main chip via chip-to-chip interconnect technology (Infinity Fabric), coupled with an independent I/O Die (IOD) that contains the memory controller, media engine, and interface modules. This modular design not only increases the core count but also provides flexibility for future expansion.
iGPU Configuration and Market Positioning #
However, as many might suspect, Medusa Point’s iGPU configuration has sparked some discussion. Compared to Strix Point’s Radeon 890M with 16 RDNA 3.5 Compute Units, the Medusa Point series’ iGPUs are reduced to 8 Compute Units, estimated to be at the Radeon 860M level. According to test data, the Radeon 860M’s graphics performance is approximately 20-25% lower than the Radeon 890M, which might struggle to match the fluidity of the previous-generation flagship in 1080p gaming scenarios. While RDNA 3.5+, an optimized version of RDNA 3.5, offers improvements in energy efficiency and specific workloads, the lack of RDNA 4 architecture support means it cannot leverage the latest ray tracing and AI acceleration technologies. This suggests that AMD’s positioning for Medusa Point prioritizes CPU performance over top-tier integrated graphics performance, likely to differentiate it from the successor to Strix Halo, which will focus on high-performance graphics processing.
Manufacturing Process and Memory Support #
Regarding the manufacturing process, Medusa Point’s Zen 6 cores will utilize TSMC’s 3nm (N3P) process, while the I/O die will use the more economical 4nm (N4P) process. The 3nm process brings higher transistor density and energy efficiency, helping control power consumption at high core counts while boosting single-thread performance. In contrast, Strix Halo’s Zen 5 cores are based on the 4nm process, providing Medusa Point’s performance uplift with a hardware foundation from the process upgrade. For memory support, Medusa Point is expected to be compatible with LPDDR5X memory, with a maximum speed possibly reaching 7500MT/s, paired with a 128-bit memory controller to further alleviate bandwidth pressure from the high core count.
Market Outlook and Competition #
From a market positioning perspective, Medusa Point targets the high-end mobile computing market, especially thin-and-light laptops and mobile workstations that require strong multi-threaded performance. Its 22-core Ryzen 9 model will particularly excel in multi-threaded workloads (such as video encoding, virtual machine execution) and is expected to directly compete with Intel’s upcoming Panther Lake series. The latter is anticipated to use the 18A process and integrate an Xe3 architecture iGPU, offering stronger graphics performance but possibly fewer cores than Medusa Point. AMD’s strategy seems to be to position Medusa Point as a CPU-performance-dominant mobile processor, attracting professional users with its core count and architectural advantages, while reserving top-tier iGPU performance for the future Medusa Halo series.
The Medusa Point series, through its Zen 6 architecture and multi-chip design, achieves a significant breakthrough in core count, bringing desktop-level multi-threaded performance to the mobile platform. This processor is positioned for professional application scenarios. The series is expected to debut in the second quarter of 2026, with CES 2026 potentially serving as its official launch stage.