Acer introduced the Veriton GN100 Artificial Intelligence mini workstation, positioning the compact system as a solution for organizations and professionals who want to run large models locally and reduce dependence on cloud services. The announcement emphasizes the device´s small footprint and its goal of lowering the operational costs associated with cloud-based model hosting and inference. The Veriton GN100 is presented as a mini workstation that brings server-class compute into a desktop-scale chassis.
The Veriton GN100 is built around the NVIDIA GB10 Grace Blackwell Superchip, which the article says delivers up to 1 PFLOPS of FP4 Artificial Intelligence performance. The platform pairs next-generation CUDA cores with fifth-generation Tensor Cores and a 20-core Arm-based CPU configuration. Memory and storage are substantial for the form factor, with 128 GB of unified system memory and up to 4 TB of NVMe M.2 SSD storage included in the configuration described. The combination of these components is framed as providing server-grade performance despite the mini-PC size.
The article focuses on the workstation´s ability to host and run large Artificial Intelligence models locally to reduce cloud reliance and cost exposure. Specific deployment scenarios, software stack, supported models, and performance benchmarks beyond the stated FP4 peak figure are not provided in the article. Similarly, details such as pricing and availability were Not stated. The announcement includes imagery and a link to a full story but does not expand on additional features, ports, or expandability options in the text presented.
