NVIDIA Jen-Hsun Huang responds to super PC processor N1: low consumption, high energy, powerful AI

NVIDIA officially launched the DGX Spark mini AI supercomputer at the end of last year, using a super SoC processor GB10. There will also be N1 series processors for desktop computers. Huang Renxun also recently talked about the progress of this chip. Huang Renxun said that they cooperated with MediaTek to create a high-energy-efficiency SoC processor.The power consumption is very low, but the energy efficiency will be very strong.

This processor is mainly designed for computer platforms and is mainly aimed at scenarios that require powerful AI applications. There are many related applications mentioned.The cooperation with MediaTek focuses on AI computers and terminal-side AI.

However, he did not mention the specific specifications and launch time of the N1 series processors.

The specifications of the N1 series are no secret. The CPU part is a 20-core ARM architecture, composed of 10 Cortex-X925 + 10 Cortex-A725.

The GPU is a Blackwell architecture, with FP4 performance reaching 1PFLOPS. It integrates the fifth-generation Tensor core and the fourth-generation RT light tracing core.48 sets of SM units, 6144 CUDA cores, and overall specifications are equivalent to RTX 5070.

Integrating an RTX 5070-level GPU into a PC processor is undoubtedly the most powerful core display at present. The core display performance of AMD and Intel's new generation processors is also inferior. However, the ultimate performance depends on the optimization of NVIDIA and MediaTek.

Judging from the previous DGX Spark that used GB10 chips, the performance of the N1 series processors cannot reach the RTX 5070 level.

The other thing is the time to market. It was announced at last year’s CES show, but we haven’t seen the product on the market for a year.The current news is that it may be launched in the Q1 quarter of this year, the first notebook platform.

However, the N1 series has been delayed so late. Next year, we can directly wait for the N2 and N2X platforms. The CPU and GPU architecture should continue to be upgraded, and the paper performance is worth looking forward to.