GenWorks is a Generative Artificial Intelligence Training Equipment platform designed for hands-on GenAI learning and development. It comes with pre-installed AI frameworks (including CUDA) and an integrated high-performance AI compute setup, delivering up to AI Performance (INT8): 3,352 AI TOPS – ideal for running and experimenting with modern generative AI workloads in education, training, and applied projects.
Product Features
- Built-in GPU based on Blackwell architecture with up to 24 cores and 32 threads CPU, and 3,352 AI TOPS class performance
- Equipped with 64GB main memory and 2TB NVMe SSD for processing large amounts of data and model parameters
- High-efficiency cooling and power design considering GPU and CPU heating for stable long-term full-load AI performance
- Minimal system package for education with Ubuntu LTS distribution, eliminating unnecessary services and simplifying GenAI environment configuration
- System image-based file system recovery mode with one-click initialization to return to the initial state anytime after practice
- Visual confirmation support for GenAI results (prompt–response process, image/video creation results) using a large 43-inch 4K monitor
- Pre-configured library and runtime environment specialized for generative AI, including Hugging Face Transformers / Diffusers, plus CUDA-based PyTorch and TensorFlow
- Pre-loaded open-source language model and diffusion model, with example projects for representative GenAI practices (text generation, code generation, image generation)
- Professional application development support, including Visual Studio Code-based IDE features, Jupyter, Git, Docker
- Generative AI educational content that can be applied directly to lectures to support a consistent curriculum connecting theory, practice, and project