Acer Launches Veriton GN100 AI Mini Workstation Powered by NVIDIA Grace Blackwell Superchip

Acer has officially unveiled the Veriton GN100 AI Mini Workstation, a compact yet exceptionally powerful desktop designed to handle large-scale AI workloads locally.

This cutting-edge system is built on the NVIDIA GB100 Grace Blackwell Superchip, delivering up to 1 petaFLOP of FP4 AI performance in a mini-PC form factor.

Engineered for AI developers, researchers, and data scientists, the Veriton GN100 is more than just a workstation, it’s a high-performance AI development hub. It combines next-gen CUDA cores, 5th-gen Tensor Cores, and 20 Arm-based CPU cores with 128 GB of unified system memory and 4 TB of NVMe M.2 SSD storage to offer server-level capabilities without the need for a dedicated data center.

The system comes preloaded with the NVIDIA AI software stack, enabling seamless integration with industry-standard tools like PyTorch, Jupyter, and Ollama. Whether you’re prototyping, fine-tuning, or deploying large language models, the GN100 empowers you to do it all locally, or scale out to the cloud when needed.

One standout feature is the NVIDIA ConnectX-7 NIC, which allows users to link two GN100 units together. This enables support for AI models with up to 405 billion parameters, making it a versatile solution for cutting-edge AI research and application development.

Key Features at a Glance: