Gigabyte launches compact AI TOP AERO mini PC with 128GB unified memory

A desktop for local AI

Gigabyte has introduced the AI TOP AERO, a compact mini PC designed specifically for running resource-heavy artificial intelligence models. The hardware aims to provide a personal AI studio experience for users who need to process complex data without relying on external servers. As mobidevices.com reported, the device is built to handle large-scale tasks while maintaining a small physical footprint.

The machine is designed to be a Windows 11 alternative to the Apple Mac Studio, focusing on high-performance computing in a small box. By providing a dedicated space for AI work, Gigabyte is targeting developers and creators who want to keep their data private. This setup allows for the local execution of models that typically require massive amounts of hardware resources.

One of the primary goals of this hardware is to enable the fine-tuning of compatible models. Because the system handles the processing locally, users do not have to send their sensitive files or model weights to the cloud. This makes it an attractive option for those who prioritize data security and want to avoid recurring cloud costs.

Superchip and memory architecture

At the heart of the AI TOP AERO is the NVIDIA RTX Spark N1X superchip. This component combines a 20-core NVIDIA Grace Arm processor with a Blackwell-based graphics processor. These two parts are connected via NVLink-C2C, which is a high-speed interconnect technology that allows them to work closely together as a single unit.

The most significant technical feature is the support for up to 128 GB of LPDDR5X unified memory. In a unified memory architecture, both the CPU and the GPU share a single pool of memory. This allows the graphics processor to directly access a large portion of the available capacity, which is essential for running models with tens of billions of parameters.

To manage the data flow, the system also includes a high-speed PCIe 5.0 SSD that can reach a capacity of 4 TB. This storage helps manage the large files and weights required for AI training and inference. The combination of high-speed memory and storage ensures that the superchip can access the data it needs without significant bottlenecks.

Performance and cooling

The hardware is capable of delivering a peak AI performance of 1 petaflop in FP4 computations. This measurement refers to the number of floating-point operations the system can perform per second. Such high performance is necessary for the rapid generation of text, images, or other AI-driven outputs in real time.

Because these components generate significant heat, Gigabyte engineers had to develop a custom cooling system. Fitting this level of power into a case with a volume of about one liter is a significant engineering challenge. The cooling solution is designed to maintain performance during sustained loads, which is common during long AI training sessions.

For connectivity, the mini PC includes standard and high-speed options like USB-C and 10GbE wired networking. It also supports Wi-Fi 7 for wireless needs. These features ensure that the device can stay connected to high-speed networks while processing the massive amounts of data required for modern AI workflows.

Software and gaming use cases

The AI TOP AERO is designed to work with several popular AI tools, including LM Studio, Ollama, and ComfyUI. It also supports NVIDIA Studio drivers, along with NVIDIA Broadcast and RTX Video technologies. These tools allow users to perform various audio and video tasks, ranging from noise cancellation to advanced video editing.

While the primary focus is on AI, the device can also run compatible games. It supports ray tracing and DLSS 5 neural rendering, which are technologies used to improve the visual quality and frame rates of modern video games. This makes the device versatile for users who want a machine that can handle both professional work and high-end entertainment.

However, there is a specific consideration for gamers regarding the chip architecture. Because the Grace processor is based on Arm, some Windows games may face compatibility issues. Users should check specific game requirements to ensure that their titles will run correctly on an Arm-based system before making a purchase.

Availability and launch

Gigabyte officially launched the AI TOP AERO on November 10, 2026. The company is positioning this as a specialized piece of hardware for the growing demand for local AI capabilities. By offering a compact form factor, they are appealing to users who want power without the bulk of a full-sized tower computer.

Pricing and specific configurations are not listed in a single universal price point. Gigabyte advises potential buyers to check regional announcements for specific details. These announcements will provide the necessary information regarding retail availability and the different hardware configurations available in different markets.

The launch marks a clear step toward making high-end AI hardware more accessible to individual creators. By combining a superchip with a large unified memory pool, Gigabyte is providing a platform that can compete with professional workstations. This move highlights the increasing importance of local hardware in the current AI landscape.