AMD Unveils Radeon 800M and Ryzen AI 300 Series: A Leap in Performance

AMD has officially confirmed that the Radeon 800M (RDNA 3.5) graphics architecture is 19% to 32% faster than its predecessor, the Radeon 700M, at 15W. Additionally, the Strix Point RDNA 3.5 architecture shows a 32% performance boost over RDNA 3 in Hawk Point at 15W in 3DMark TimeSpy.

Benchmark Revelations

AMD has released official benchmark scores, showcasing the performance of the Radeon 800M and Radeon 700M graphics in 3DMark TimeSpy and Night Raid benchmarks. The Radeon 800M architecture demonstrates a 32% improvement in 3DMark TimeSpy and a 19% speed boost in Night Raid.

Introducing the Ryzen AI 300 Series

Set to launch later this month, AMD’s Ryzen AI 300 series, codenamed “Strix Point,” brings significant upgrades, including Zen5 CPUs, XDNA2 NPUs, and the new RDNA 3.5 graphics architecture. This new architecture focuses on enhancing power-performance efficiency, improving performance per watt for common graphics shader operations, and reducing memory access requirements to deliver better battery life.

Key Features of RDNA 3.5

2x Texture Sampler Rate: Common texture sampling operations are now double rate for game textures.
2x Interpolation and Comparison Rates: Enhanced vector ISA for interpolation and comparison in shaders.
Improved Memory Management: Better primitive batch processing, compression techniques, and workload reduction.

Competitive Edge

In gaming benchmarks, AMD’s focus extends beyond comparing RDNA 3 to RDNA 3.5. The company aims to compete with Intel’s Core Ultra 9 185H, featuring 8 Xe-Cores. AMD claims RDNA 3.5 will deliver performance gains ranging from 27% to 65% depending on the game.

Snapdragon Elite APU Performance

AMD also tested the Snapdragon Elite APU, which faced issues in 3 out of 7 tests and showed slower performance compared to both Intel and AMD solutions.

Launch Date

The Ryzen AI 300 series is expected to launch on July 28th.

Stay tuned for more updates as AMD continues to push the boundaries of performance and efficiency in the graphics and AI processing landscape.