AMD medusa point chip appears in geekbench

An unannounced AMD Medusa Point mobile processor has surfaced in Geekbench, offering an early look at a Zen 6-based design with 10 cores and a larger cache configuration. The listing points to active pre-release testing, but the benchmark results remain too immature to judge performance.

A Geekbench listing has revealed an unannounced AMD processor carrying the OPN code 100-000001713-31. The platform is identified as Plum-MDS1, which points to Medusa Point, AMD’s next-generation mobile APU family based on Zen 6. The chip is a 10-core, 20-thread part with a 2.4 GHz base clock, although the recorded test behavior shows it running closer to 1.3-2 GHz, consistent with an engineering sample still early in development.

Cache details in the listing suggest a notable step up for this class of mobile chip. Each core gets 1 MB of L2 cache, and the L3 comes in at 32 MB, up from 24 MB on Strix Point and 16 MB on Hawk Point. This is 50% more cache than the current 10-core parts like the Ryzen Artificial Intelligence 9 365. The test system had 32 GB of memory installed, though the memory type was not specified.

The benchmark scores themselves offer little insight at this stage. The chip spent most of the test hovering around 1.39 GHz and barely peeked above 2 GHz, well below what a final retail part on a 3 nm process would run at. That makes the current results too preliminary to support performance conclusions.

Separate shipping manifests from Planet3DNow associated with the Medusa codename suggest a 4C+4D layout with standard and density-optimized cores, though that does not fully explain the 10-core count shown in Geekbench. One possible explanation is that two additional low-power cores are placed in the IO die, bringing the total to ten. The processor is expected to be a 28 W TDP mobile chip for the FP10 socket, and Medusa Point is described as combining Zen 6 CPU cores with RDNA 5 and RDNA 3.5 graphics, along with an updated NPU. A launch around CES 2027 would fit AMD’s usual cadence, but the Geekbench entry indicates that testing is already underway.

52

Impact Score

Mistral launches text-to-speech model

Mistral has expanded its Voxtral family with a text-to-speech system aimed at enterprise voice applications. The company is positioning the open-weights model as a flexible alternative for organizations that want more control over deployment, cost and customization.

UK Parliament opens workforce inquiry on Artificial Intelligence

A UK Parliament committee is examining how Artificial Intelligence is changing business and work, with a focus on both economic opportunity and labour disruption. The inquiry is seeking evidence on government priorities as adoption expands across the economy.

Windows 11 tightens kernel trust for older drivers

Microsoft is changing Windows 11 kernel policy so new drivers must be signed through the Windows Hardware Compatibility Program. Older trusted drivers will still be allowed in some cases to preserve compatibility during the transition.

Proofpoint expands security for the agentic workspace

Proofpoint introduced new email and data security capabilities designed for workplaces where humans and Artificial Intelligence agents interact across communication and data environments. The updates combine email protection models, add data access governance for human and non-human identities, and extend data security posture management into on-premises systems.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.