It turns out that there was more to yesterday’s press event regarding the new partnership, as Jensen Huang also revealed more about its ARM-based processor. It’s no secret that NVIDIA is working on an ARM-based processor called N1, which was aimed at laptops and PC; it was not known that N1 had a much larger role on the horizon for enterprise/datacenter applications.
As reported by VideoCardz:
“We’re building arm, of course, robotics processors. our latest one is called thor. it’s used for robots and of course, for autonomous driving. we also have a new arm product that’s called n1 and that product is, that processor is going to go into the Dgx spark and many other versions of products like that, and so we’re super excited about the arm roadmap, and this doesn’t affect any of that.”
-Jensen Huang, NVIDIA CEO
One of the reasons that the GB10 is referred to as a Superchip is that the package contains a CPU capable of up to 1 petaFLOPs with FP4 precision along with a Blackwell-based graphics solution. More about the NVIDIA DGX Spark was revealed at GTC25 (video here), and then details about GB10 were shared at Hot Chips 2025 (via ServeTheHome). Here’s a quick overview of the DGX/GB10 specs.
- Arm CPU made in collaboration with Mediatek
- Capable of 1 petaFLOP AI performance (per DGX Spark official page)
- Built on TSMC 3nm process
- 20 Arm v9.2 cores divided into 2x clusters of 10
- 16 MB L3 cache split among each cluster
- LDDR5 256-bit memory supporting up to 128 GB configuration at 301 GB/s
- Package TDP up to 140 W
- Supports multiple displays, including HDMI 2.1 and DP Alt-mode
“Powered by the NVIDIA GB10 Grace Blackwell Superchip, NVIDIA DGX™ Spark delivers 1 petaFLOP of AI performance in a power-efficient, compact form factor. With the NVIDIA AI software stack preinstalled and 128GB of memory, developers can prototype, fine-tune, and inference the latest generation of reasoning AI models from DeepSeek, Meta, Google, and others with up to 200 billion parameters locally, and seamlessly deploy to the data center or cloud.”
-NVIDIA
Having already conquered the consumer and enterprise GPU sector, it’s clear that NVIDIA now has a new strategy centered on the CPU ecosystem. Between x86 (via its new partnership with Intel) and Arm (via its partnership with Mediatek), the chess pieces are on the board.