AMD‘s next processor architecture, known by the codename Medusa Point (Zen 6), is moving forward faster than expected and is already showing stronger performance than earlier estimates.
New benchmark results found in the Geekbench database reveal that an engineering sample of the Medusa Point processor clearly outperforms its predecessor based on the Strix Point (Zen 5) architecture.
The leaked chip, listed under the AMD Plum-MDS1 platform, features 10 cores and 20 threads. In Geekbench 6.6.0 testing, it reached a maximum recorded clock speed of 5.37 GHz, or about 5.4 GHz. That is around 300 MHz faster than the top Strix Point models.
AMD Medusa Point Engineering Sample Shows Improvement Over Zen 5
The benchmark results also show noticeable improvements over Zen 5.
In single-core testing, the processor scored 3,329 points, which is about 5% higher than previous leaks of the same engineering sample. Compared with the current Ryzen AI 9 365 with 10 Zen 5 cores, the new Medusa Point chip delivers almost 35% better single-core performance.
In multi-core testing, it achieved 16,555 points, representing a 9% improvement over benchmark results that appeared a few weeks ago. It also performed well above the Ryzen AI 9 365, which averages 12,407 points, and even surpassed the 12-core Ryzen AI 9 HX 370, which scores around 13,396 points.
While synthetic benchmarks such as Geekbench should be treated as early indicators rather than final performance figures, these results suggest that Zen 6 offers a significant increase in IPC performance along with improved efficiency.
AMD still has several months before the official launch of the Medusa Point family. During that time, engineers are expected to continue refining clock speeds and power consumption. Even so, these early engineering samples suggest AMD is aiming for a major generational leap and hopes to strengthen its position in the market for high-efficiency and portable APUs.
