It's known that AMD's Zen 4 processors is one of the best CPUs around. Phoronix answers those who wonder if the mobile Zen 4 can compete with rivaling mobile chips such as Apple's M2 silicon. The Linux publication has compared two mobile Zen 4 processors against the Apple M2 under several Linux benchmarks.
The Ryzen Z1 Extreme is a high-performance Zen 4 processor that powers handheld gaming devices like the ROG Ally. The octa-core, 16-thread chip has a 5.1 GHz boost clock and a cTDP between 9W to 30W. The Ryzen 7 7840U, which resides inside Acer's Swift Edge 16, is another processor from AMD's Zen 4 camp, wielding a similar octa-core, 16-thread design with a 5.1 GHz boost clock. Unlike the Ryzen Z1 Extreme, the Ryzen 7 7840U has 28W, although the chip has a cTDP between 15W and 30W.
On the contrary, the M2, which powers the MacBook Air 2022, features an octa-core, Arm-based design with four performance "Avalanche" cores and four efficiency "Blizzard" cores. The Avalanche cores max out at 3.5 GHz, whereas the Blizzard cores can reach 2.4 GHz. It's worth remembering that the MacBook Air features a passive cooling solution, unlike the MacBook Pro. The ROG Ally and Swift Edge 16 come with active cooling solutions. More importantly, Phoronix tested the 8GB version of the MacBook Air 2022. The news outlet used Asahi Linux for testing the MacBook Air 2022, and while support is commendable, the project is pretty much a work in progress. Therefore, the M2 likely has untapped potential due to the lack of proper Linux support.
According to the geometric of the test results, AMD's Zen 4 processors wiped the floor with the Apple M2. With the balanced mode on the ROG Ally, the Ryzen Z1 Extreme delivered 28.7% higher performance than the Apple M2. The performance delta increased to 95.7% when Phoronix put the ROG Ally into performance mode. Meanwhile, the Ryzen 7 7840U outperformed the Apple M2 by 75.8%.
The Ryzen Z1 Extreme and Ryzen 7 7840U showed good performance-per-watt in the benchmarks. Sadly, a PowerCap/RAPL or HWMON driver for the Apple M2 doesn't exist, so Phoronix could not log the silicon's real-time power consumption metrics with Linux. Many would agree it would be interesting to see the Apple M2's power efficiency compared to the Zen 4. True, Phoronix could evaluate the AC power from the wall, but it doesn't provide a precise measurement.
Phoronix's testing satisfies the curiosity of those who want to see a fight between Zen 4 vs. Apple M2. Some may argue that it's not an apples-to-apples comparison since there are too many variables, such as the difference in cooling and that the Apple M2 was running on Asahi Linux instead of the native macOS, which is better optimized for the M2 silicon.
