AMD Medusa Point Zen 6 APU leak points to impressive early performance at just 2 GHz
AMD’s next-generation Medusa Point APU has appeared again in the Geekbench database, and the latest results suggest that the upcoming Zen 6 architecture could deliver a major performance uplift when it arrives in future laptops and compact PCs.
The newly spotted chip is listed as a Ryzen 9 processor with the CPU ID “100-000001713-33_N.” It features 10 cores and 20 threads, reportedly built on AMD’s Zen 6 CPU architecture. The core layout appears to use a 4+6 configuration, which likely combines standard Zen 6 cores with efficiency-focused variants designed to improve performance per watt.
This is not the first time Medusa Point has surfaced in early benchmark listings. A previous entry showed the chip running on AMD’s Plum-MDS1 evaluation platform, which uses the new FP10 BGA socket. That platform is believed to be intended for next-generation AMD system-on-chips in the 28W to 45W power range, making Medusa Point a likely candidate for premium notebooks, mini PCs, and other power-efficient systems.
What makes the latest benchmark particularly interesting is the clock speed. The newly tested Medusa Point APU was listed with a 2.00 GHz base clock and a maximum clock of around 2.06 GHz. Despite these very early and relatively low frequencies, the chip managed to deliver strong benchmark results.
In Geekbench, the 10-core AMD Medusa Point Zen 6 APU scored 3,174 points in single-core performance and 15,092 points in multi-core performance. Those numbers are notable because the chip is still running at much lower clock speeds than current-generation Ryzen AI processors based on Zen 5.
For comparison, AMD’s Ryzen AI 9 365, a 10-core Zen 5 “Strix Point” chip, can boost up to around 5 GHz. Even with that large clock-speed advantage, the early Zen 6 engineering sample appears highly competitive. Based on average Geekbench results, the Ryzen AI 9 365 remains ahead by roughly 29% in single-core performance and about 22% in multi-core performance, but it is doing so at more than double the operating frequency.
That gap may narrow significantly once final Zen 6 silicon reaches production-level clock speeds. Engineering samples often run at conservative frequencies, and early firmware or platform limitations can prevent chips from showing their full potential. If AMD can scale Medusa Point toward higher boost clocks while maintaining efficiency, Zen 6 could represent a substantial generational jump.
The leaked listing also shows 10 MB of L2 cache and 32 MB of L3 cache, although the L3 cache value may not be reported correctly by Geekbench at this stage. Early benchmark databases often misread cache details for unreleased processors, so final specifications may differ.
Another key detail in the listing is support for FP16 AVX-VNNI instructions. This suggests that AMD may be preparing expanded FP16 capabilities for Zen 6-based SoCs. FP16 support can help accelerate workloads involving artificial intelligence, machine learning, media processing, and other compute-heavy tasks that benefit from lower-precision operations.
That addition would make sense for Medusa Point, as AMD has been pushing heavily into AI PC hardware with its Ryzen AI lineup. Future Zen 6 APUs are expected to combine stronger CPU cores, improved integrated graphics, and enhanced AI acceleration, creating a more complete platform for next-generation mobile computing.
Medusa Point is currently expected to launch around 2027, with more details likely to emerge closer to CES 2027. While specifications and performance figures may change before release, these early Geekbench results offer an encouraging look at AMD’s next major CPU architecture.
If the current numbers are any indication, Zen 6 could bring higher efficiency, stronger per-clock performance, improved AI-related instruction support, and a broader performance uplift across both single-core and multi-core workloads. For users waiting to upgrade their laptops or compact PCs, AMD’s Medusa Point APUs may become one of the most important processor launches to watch in the coming years.
