Nvidia N1x ARM Performance Lags Behind Apple M3 Max in Geekbench

Nvidia is reportedly preparing a major announcement at Computex: a new ARM-based platform called N1x, developed in collaboration with Microsoft and Arm for future laptops. However, early pre-release Geekbench 6 results have already sparked skepticism, as the new Nvidia chip apparently cannot outperform the Apple M3 Max, which debuted in the 2023 MacBook Pro.

The N1x is believed to be a modified version of the GB10 system-on-chip used in the DGX Spark mini-PC. The platform is rumored to feature a 20-core ARM processor developed by MediaTek, graphics on par with an RTX 5070, and unified LPDDR5X memory shared between CPU and GPU. This architecture should be particularly appealing for local AI tasks, where large memory capacity and tight integration of compute units matter.

In a discovered Geekbench 6.2.2 test for Linux AArch64, the Nvidia N1x scored 3,096 points in single-core and 18,837 points in multi-core. The system ran on Ubuntu 24.04.1 LTS with an HP 8EA3 motherboard. For comparison, the Apple M3 Max in a 16-inch MacBook Pro achieves around 3,124 points in single-core and 18,920 points in multi-core. The difference is small, but the Apple chip still holds a formal advantage.

The comparison looks especially awkward for Nvidia given the configuration gap. The M3 Max uses a 14-core CPU, while the N1x is rumored to have a 20-core processor block. If final figures remain close to these values, it would highlight not only Apple's optimization strength but also that Nvidia and its partners still need to prove the new ARM platform's efficiency in consumer devices.

That said, it's too early to draw definitive conclusions. The N1x tests, as noted, were from June 2025 and belong to a pre-release platform. Production versions in optimized Windows laptops could show higher results, especially if Microsoft and Nvidia refine drivers, power profiles, and application support. Moreover, Geekbench primarily evaluates CPU performance, whereas the N1x's strong suit may be graphics and local AI workloads rather than traditional processor performance.

The big question for Computex is how Nvidia will position the N1x. If the company presents it as a direct universal competitor to Apple Silicon, comparisons with the M3 Max will be inevitable and not entirely flattering. But if the focus is on a new type of PC where GPU, unified memory, and local models take precedence, then these early CPU benchmarks may only tell part of a more complex story.