Search
This week, at the GTC Taipei event, Nvidia officially introduced the RTX Spark superchip, designed specifically for personal AI computers. The processor segment of this chip was co-developed by Nvidia and MediaTek, leveraging the Arm architecture. Meanwhile, the GPU segment incorporates Nvidia's Blackwell architecture, integrating a substantial 6,144 CUDA cores. This powerhouse can deliver an impressive AI computing performance of up to 1 petaFLOP, supports unified memory configurations of up to 128GB, and is capable of running AI models with up to 120 billion parameters locally. Starting this fall, leading PC manufacturers including Lenovo, Dell, and HP are set to roll out a series of thin-and-light laptops and compact desktops, all equipped with the RTX Spark chip.
-
C114 Communication Network -
Communication Home
