AMD Zen 6 "Medusa Point" Surfaces in Geekbench 7 Leak: Outscores Ryzen 9 9950X3D in Single-Core
AMD Zen 6 "Medusa Point" Surfaces in Geekbench 7 Leak: Outscores Ryzen 9 9950X3D in Single-Core
An early 10-core engineering sample running on a Lenovo platform hints at robust architectural uplifts for AMD's upcoming next-gen mobile lineup.
First architectural details regarding AMD’s upcoming **Zen 6 "Medusa Point"** processors have emerged via a Geekbench 7 leak. Tested on a Lenovo validation platform, the 10-core engineering sample posted impressive numbers, notably eclipsing high-end desktop gaming chips like the Ryzen 9 9950X3D in single-core workloads.
As the industry transitions toward next-generation mobile and ultra-portable platforms, these preliminary benchmark metrics provide a fascinating glimpse into the performance potential of AMD's upcoming Zen 6 microarchitecture.
1. Benchmark Breakdown: Single-Core & Multi-Core Performance
The leaked engineering sample (identified as 100-000001713-32_Y) scored 3,100 points in the Geekbench 7 single-core suite and 18,313 points in multi-core testing.
Geekbench 7 Single-Core Comparison
Zen 6 Medusa Point
3,100 pts
10-Core ES Sample
Ryzen 9 9950X3D
3,051 pts
Desktop Flagship
Ryzen 9 9950X
3,069 pts
Standard Desktop
Ryzen 9 9900X
3,023 pts
12-Core Desktop
The single-core score puts the early Zen 6 silicon roughly 1.6% ahead of the Ryzen 9 9950X3D, signaling substantial IPC (instructions per cycle) and architectural efficiency gains even before final clock speed optimization and retail power limits are locked in.
2. Multi-Core Muscle & Cluster Configuration
In multi-threaded evaluation, the processor achieved 18,313 points—outperforming the 12-core Ryzen AI MAX 390 (17,393 points) and landing just 1.2% behind the Ryzen AI MAX PRO 390 despite sporting fewer physical cores.
| Specification / Feature | Observed Engineering Sample Details |
|---|---|
| Architecture & Lineup | Zen 6 "Medusa Point" (Mobile / Next-Gen APU) |
| Core / Thread Count | 10 Cores / 20 Threads (Configured in 4+6 clusters) |
| L3 Cache Allocation | 16 MB L3 Cache (Reporting varies across engineering revisions) |
| Platform & Clocks | Lenovo test bench; early base clock ~2.0 GHz (ES state) |
The 10-core layout is arranged across asymmetric clusters (four cores in one block and six in another), demonstrating AMD's ongoing refinement of power-efficient multi-core scheduling for mobile form factors.
Outlook for the Zen 6 Generation
While engineering samples running at early clock speeds do not represent final retail performance, the ability of a mobile-oriented 10-core Zen 6 chip to challenge or beat flagship Zen 5 desktop processors in single-threaded tasks bodes exceptionally well for AMD's next-gen mobile lineup. Official specifications, final clock targets, and commercial release windows remain unconfirmed by AMD.