The global PC market is undergoing an unprecedented reshaping. Recently, NVIDIA's official social media account posted a message containing only one sentence and a string of mysterious numbers—“PC New Era. 25.0528, 121.5990”—which immediately sparked a frenzy in the tech community. Almost simultaneously, Microsoft and ARM also released the same update. The synchronized moves by these three giants clearly point to the long-anticipated NVIDIA self-developed PC chip, signaling that a technological battle between the Windows on Arm alliance and Apple's M-series chips is about to begin.

The mysterious set of geographic coordinates directly pinpointed the Taipei Pop Music Center in Taiwan, which is the main venue for the upcoming Computex (Taipei International Computer Show). At that time, NVIDIA CEO Jensen Huang will deliver a keynote speech, and the core focus of this speech is likely to be the highly anticipated NVIDIA self-developed PC chip—N1X.

image.png

From the currently leaked technical specifications, the N1X can be considered NVIDIA's "M-series super chip." According to reports, this chip was jointly developed by NVIDIA and MediaTek, based on TSMC's most advanced N3B manufacturing process. It not only features a 20-core ARM architecture CPU but also integrates a Blackwell architecture GPU with 6144 CUDA computing units. In terms of memory design, it adopts a unified memory architecture similar to Apple's, equipped with 128GB LPDDR5X memory, allowing both the CPU and GPU to share access. Just looking at the number of 6144 CUDA cores, its graphics computing power is already comparable to traditional desktop discrete GPUs.

image.png

However, this revolutionary chip also has some practical constraints. Due to the shared LPDDR5X unified memory between the CPU and GPU, the bandwidth is expected to be around 273 GB/s, and running traditional x86 games on the ARM architecture requires an additional layer of emulation translation. This means that if users expect to use it as a pure ultrathin gaming laptop, the actual experience may be compromised.

But in the era of AI-native, gaming performance seems no longer the core mission of this machine. For developers and creators, the N1X is more like a local "compute printer." In the past, ordinary people had to bear high cloud subscription fees and token consumption costs to access cutting-edge large models; the emergence of high-performance AI-native PC devices like NVIDIA allows models to be deployed and run locally. Users only need to bear a one-time hardware investment to freely produce tokens at almost zero marginal cost, completing automated tasks, hot topic extraction, or file organization. This compute decentralization is breaking the monopoly of big companies on AI infrastructure, enabling every individual's creativity to become a reality without restraint.