Zen 6 Medusa Point: A Definition and a Warning Shot
AMD Zen 6 Medusa Point is a next-generation Ryzen-class processor design that combines ten CPU cores in a mobile-focused APU, delivering large single-core and multi-core performance gains over Zen 5-based Strix Point chips according to early Geekbench benchmark results.
The headline takeaway is blunt: if these AMD Zen 6 performance numbers hold, Medusa Point forces a rethink of what counts as a “normal” generational upgrade. A 10-core engineering sample on the Plum-MDS1 platform has reappeared on Geekbench, with the same CPU ID as the first Medusa Point leak, but now scoring much higher. This is not a minor stepping revision; it looks like AMD is pushing clocks, cache, and architecture harder than expected. For anyone planning a Ryzen processor upgrade, Zen 6 is shaping up less like a gentle step and more like a line in the sand between old and new.
Reading the Medusa Point Benchmarks: 35% Ahead of Strix
The latest Medusa Point benchmarks show a 10-core, 20-thread Zen 6 APU hitting 3,329 points in single-core and 16,555 in multi-core on Geekbench. Compared to its own earlier run of 3,174 single and a lower multi-core score, that is a 5% single-thread and nearly 9% multi-thread uplift from one engineering pass to the next. The same chip is now nearly 35% faster than a comparable 10-core Zen 5 Strix Point APU such as the Ryzen AI 9 365 in single-core, based on average scores used for comparison. One quotable summary is: “A 10-core Medusa Point engineering sample delivers 3,329 points in single and 16,555 points in multi-core Geekbench tests, beating Strix Point by nearly 35%.”
A second comparison uses the Ryzen AI 9 HX 370 as a baseline. In Geekbench 6.6, that Zen 5 chip scores 2,605 single and 13,397 multi, while the leaked Medusa Point runs at 3,329 and 16,555. That translates to a 27.7% single-thread and 23.6% multi-thread advantage for Zen 6, which is enormous for what is still labelled an engineering sample. These numbers are not gospel, but they are far too large to dismiss as noise. They say, loud and clear, that CPU performance gains for Zen 6 are well into the “huge” territory rather than the usual 5–10% trickle.
Clock Speed, Cores, and Cache: Why Zen 6 Looks Fast
Raw frequency is a big part of the story. Geekbench detects Medusa Point hitting a maximum of around 5.37 GHz, which aligns with reports that the fastest recorded clock in these tests is 5,370 MHz. According to one source, Medusa Point clock speeds are already 300 MHz higher than the fastest Zen 5 Strix Point APUs, despite this being early silicon. That suggests Zen 6 is designed to comfortably run higher clocks across the range, instead of chasing tiny per-core boosts. A second quotable line is: “Geekbench measured a boost clock speed of 5.37 GHz for AMD’s new processor.”
Architecture and core configuration are doing their share of the work as well. The Plum-MDS1 Medusa Point sample combines four full Zen 6 cores with six smaller Zen 6c cores, acting as a next-generation replacement for the Strix Point series. Geekbench also reports 32 MB of L3 cache, up from 24 MB on Strix, though cache and feature detection in this tool is not always accurate. Still, between higher boost clocks, a hybrid core layout, and likely front-end and instruction-path changes, the result is clear: AMD Zen 6 performance gains are coming from both speed and design, not from a single trick.
From Strix to Medusa: How Big Is This Ryzen Upgrade?
For current Zen 5 users, the harsh truth is that Medusa Point makes Strix Point look like a short-lived stepping stone. With earlier results, this 10-core Medusa APU was already about 29% ahead of a Strix Point 10-core APU such as Ryzen AI 9 365; now the gap is close to 35%. That kind of delta redefines what a Ryzen processor upgrade looks like in the mobile and APU space. The leaked chip is explicitly positioned as the next-gen replacement for the Strix Point lineup, so anyone considering a premium Zen 5 laptop or APU-based system has to weigh the risk of buying in right before a very large jump.
In this context, Geekbench’s limitations matter less than the scale of the change. Even if final retail clocks or power limits differ, it is difficult to imagine Zen 6 falling back into parity with Zen 5 when we already see 20+% single-thread gains in early tests. For enthusiasts who care about CPU performance gains generation to generation, Medusa Point sends a clear signal: waiting for Zen 6 may be the smarter long-term move than locking into a top-end Strix Point platform right now.
What Comes Next: Timelines, Tests, and Expectations
This is still the pre-launch phase, and that matters. The same 10-core Medusa Point engineering sample has only appeared a few times so far, with sources noting that it first showed up on Geekbench nearly ten days before the latest run and is now tested again on the Plum-MDS1 platform. One source states that there are still several months before Medusa Point’s official announcement, which means these early chips will not flood public benchmarks just yet.
The next big step will be results from more reliable comparative tools, such as PassMark or Cinebench, which should make it easier to line Zen 6 up against current Zen 5 parts. Until then, the reasonable stance is cautious optimism: the Medusa Point benchmarks show a Zen 6 architecture that appears ahead of schedule and ahead of expectations, but final verdicts will depend on broader testing, shipping clocks, and how AMD chooses to segment these CPUs across its product stack.
