Gigabyte introduced the AI Top Aero mini PC featuring the Nvidia RTX Spark platform with 128 GB of unified memory, Jagat Review reported.
The device packs an RTX Spark N1X Superchip containing a 20-core Grace processor and a Blackwell GPU with 6,144 CUDA Core units. Storage configurations reach up to 4 TB using PCIe 5.0 solid-state drives, while the system architecture is built to run large local artificial intelligence models without cloud reliance. For connectivity, tersedia USB Type-C, Ethernet 10 GbE, and Wi-Fi 7.
Personal AI Studio Configuration
NVIDIA stated that the underlying platform delivers computing performance scaling up to 1 petaflop in FP4 format. Developers can use software packages such as LM Studio, Ollama, and NVIDIA Nemotron directly on the hardware.
“Gigabyte AI Top Aero ditenagai RTX Spark N1X Superchip,” Jagat Review reported regarding the processing specifications. The system also supports NVIDIA PAIR to distribute computational tasks across multiple local machines.
Connectivity options and Samsung Galaxy A27 5G pricing
Physical connections include USB Type-C ports, 10 GbE Ethernet, and Wi-Fi 7 adapters. Local pricing details remain unannounced by the manufacturer.
Earlier hardware iterations utilizing the DGX Spark platform with 64 GB of RAM and a 2 TB solid-state drive debuted at 4,999 United States dollars, indicating a higher retail entry point for the new RTX Spark configurations.
What processor and memory powers the Gigabyte AI Top Aero?
What processor powers the Gigabyte AI Top Aero?
The system runs on an Nvidia RTX Spark N1X Superchip featuring a 20-core Grace processor paired with a Blackwell graphics processing unit.
What is the maximum memory capacity available?
Users can configure the mini PC with up to 128 GB of LPDDR5X unified memory.
Which local AI software applications are supported?
Supported applications include LM Studio, Ollama, and NVIDIA Nemotron models.
What storage options does the mini PC accommodate?
The device supports PCIe 5.0 solid-state drives with capacities reaching up to 4 TB.
Keep reading