AMD deep-dives Zen 5 architecture — Ryzen 9000 and AI 300 benchmarks, RDNA 3.5 GPU, XDNA 2, and more (2024)

AMD Zen 5 Ryzen AI 300 ‘Strix Point’ Gaming and Productivity Benchmarks

• Page 1: Zen 5 Barrels into the Market
• Page 2: Power Efficiency, Thermals, Chipsets and Overclocking
• Page 3: AMD Zen 5 Ryzen AI 300 ‘Strix Point’ Gaming and Productivity Benchmarks
• Page 4: Zen 5 Microarchitecture
• Page 5: RDNA 3.5, XDNA 2 engine, and Thoughts

AMD also unveiled the Ryzen AI 300 ‘Strix Point’ processors at Computex 2024. The first laptops with the processors will be available this month at major retailers. AMD didn’t announce any new Strix Point chips but shared new benchmarks.

AMD will launch with the 12-core, 24-thread Ryzen AI 9 HX 370 (four Zen 5 cores and eight density-optimized Zen 5c cores) and the 10-core, 20-thread Ryzen AI 9 365 (four Zen 5, six Zen 5C).

The company’s benchmarks focused solely on the Ryzen AI 9 HX 370 in the Asus Zenbook S 16 compared to the Intel Core Ultra 9 185H in the Asus Vivobook S 16, and, of course, the elephant in the room: The Snapdragon X Elite X1E-84-100 in the Samsung Galaxy Book 4 (test notes are at the end of the album).

AMD’s productivity benchmarks show a 7% advantage in PCMark 10’s application startup test, a 13% lead in the full PCMark 10 benchmark, a 21% lead in the Procyon office productivity suite, and a 30% advantage in the Kraken web browsing JavaScript benchmark. However, these percentages appear to be relative to the Intel Core Ultra chip and not the Snapdragon laptop. The lack of a labeled axis makes comparisons to the X Elite difficult.

We see another round of somewhat hard-to-decipher results in the content creation tests, but the Ryzen processor takes the overall lead in all but the single-threaded Cinebench 2024 benchmark. There, the Snapdragon X Elite takes the lead, despite the 1.2X label on the AMD chip, and the X Elite is also within striking distance in the multi-threaded Cinebench benchmark. AMD enjoys a massive advantage in the Blender benchmark and carves out respectable leads in HandBrake and PCMark 10, too.

AMD also took on its competitors in gaming, claiming anywhere from a 27% to 65% lead over the Intel Core Ultra 9 processor. Still, again, the specific advantages over the X Elite are less discernible due to poor chart labeling. AMD is however quick to point out that three of the game titles failed to run on the X Elite processor, which isn’t entirely surprising.

We always expect inconsistencies in vendor-provided benchmarks, so take all these results with a grain of salt. The test notes (end of the album) also indicate that the Intel processor only had 16GB of memory compared to the 32GB used for the Ryzen processor, raising the question of whether the Intel system was running in a performance-killing single-channel config — hopefully not, but the lack of a consistent configuration on the memory certainly isn't great to see.

AMD has traditionally shied away from direct performance comparisons with Apple’s M series processors. However, Asus presented its new AMD-powered Zenbook S 16 at the event and shared some of its own benchmarks to highlight the Ryzen AI 9 HX 370’s performance against a competing Apple MacBook Air 15 with an M3 processor. Asus provided sparse test config info in the slide, so we’ll have to take these benchmarks with more than the usual amount of salt.

Asus claims substantial leads over the MacBook Air 15, with wins ranging from a 20% advantage in the Geekbench OpenCL CPU score benchmark to a whopping 118% lead in the UL Procyon benchmark. Other notable wins include a 60% advantage in Cinebench (certainly the multi-threaded benchmark) and a 20% lead in the Geekbench CPU score.

Asus also touted the performance gains to be had with AMD’s Fluid Motion Frames (AFMF) and the 54W Strix Point’s Radeon 890M integrated graphics, showing the combination delivering more performance than a discrete RTX 2050 GPU (60W GPU + 45W CPU). As you would expect, Radeon only led because of the extra frames added by AFMF, a driver-level implementation of frame generation that doesn't always work as well as charts might indicate. (We have two copies of that slide in the deck above, one with the Radeon scores without AFMF and the comparative slide with AFMF enabled.)

Paul Alcorn

Managing Editor: News and Emerging Tech

Paul Alcorn is the Managing Editor: News and Emerging Tech for Tom's Hardware US. He also writes news and reviews on CPUs, storage, and enterprise hardware.