AMD Ryzen 9 5900HX / ASUS ROG Strix G15 AMD Advantage On Linux

Those wishing to see how their own Linux laptop(s)/system(s) compare to the ASUS ROG Strix G15 with Ryzen 9 5900HX, running your own fully-automated side-by-side comparison can be achieved by installing the Phoronix Test Suite and running phoronix-test-suite benchmark 2107215-IB-5900HXLIN46 for your own side-by-side comparison against all of the results found in this article plus all 158 tests in full.

Some additional tests on the ASUS ROG Strix G15 were also carried out looking at the additional performance characteristics.

When looking at the CPU peak frequency found on any of the cores, the Ryzen 9 5900HX was indeed hitting its boost clocks fine. Across the mix of single and multi-threaded workloads, the peak frequency observed on the 5900HX tended to be about 4.1GHz.

The CPU power consumption across a mix of workloads was 56 Watts based on the exposed RAPL information by the kernel. A peak of 153 Watts was reported.

As mentioned, unfortunately there is no CPU temperature support with the mainline kernel until Linux 5.15. The average system temperature reported by the G15 was 69 degrees under load with a peak of 90 degrees. See more of these thermal/power metrics via this OpenBenchmarking.org result file.

Long story short, the AMD Ryzen 9 5900HX (and Radeon RX 6800M, albeit to be covered in a separate article) perform very well under Linux. However, there are headaches involved at least as far as the popular ASUS ROG Strix G15 is concerned... Existing Linux distributions don't play nicely out-of-the-box until on at least Linux 5.13. It's unfortunate in 2021 that a modern laptop can have issues as fundamental as keyboard input or working WiFi, but alas that's the case here. At least Linux 5.13 otherwise is largely working out well but needing Linux 5.14 for suspend/resume and 5.15 down the road for k10temp CPU temperature reporting.

Stay tuned for more Ryzen 9 5900HX / Radeon RX 6800M Linux benchmarks from the ROG Strix G15 over the weeks ahead.