Taiwan AI Cloud Products Lines Empower Generative AI with AIHPC
From the computing layer to the platform layer, we create enterprise-level generative AI services and platforms with open-source models, provides users with LLM democratization, drives trustworthy AI, affordable AI, and sustainable AI, and empowers enterprise AI 2.0 application innovation and deployment expansion.
Taiwan AI Cloud Provides Open-Source LLMs Democratization
AI 2.0, we are committed to promoting the commercialization and democratization of trustworthy AI.
Trustworthy AI
We use a secure and reliable AIHPC supercomputer environment to create enterprise-level generative AI services and platforms for various trusted open-source models. Users have 100% control over large language model selection, runtime environment, and content fine-tuning, and also have multiple options in development tools, large language models, fine-tuning methods, deployment methods, and inference usage.
Affordable AI
We have integrated multiple models into four major generative AI development products and services, and we constantly update and optimize software and hardware environments and functions to help users reduce costs and barriers. We are committed to creating AI technology for the public, and helping enterprises innovate in generative AI applications and services to achieve AI for All – Smart Technology for People.
Sustainable AI
Building Enterprise-level Generative AI Fast for Your Business
AI 2.0, LLMs Driving the New Transformation of the AI Industry Chain
Computing power x Big data x LLM Explosive Demand
#AIHPC Faster for your GenAI Business
Taiwania 2 AI Supercomputer provides AIHPC high-performance computing and cloud platform services with GPU as the core.
With abundant computing power, an efficient parallel computing environment, and large language model-cutting technology, enterprise AI 2.0 application development is made faster and easier!
Three Major Services Empower Enterprises with AI Autonomy
Providing more diverse choices for various scenarios and applications, including development and training tools, open-source models, model fine-tuning, deployment, and inference methods.
AFS
HPC
OneAI
AFS → Workbench For Model Optimization & Deployment One-stop solution for creating enterprise-specific large language models, which can be trained, adjusted, and built into proprietary large models anytime.
A one-stop solution specifically designed to assist enterprises in creating their proprietary large language models. Enterprises can train, adjust and build exclusive large language models at any time. The solution offers various models, including FFM-Llama2, FFM-BLOOM, Embedding, with enhanced Traditional Chinese text data. Enterprises only need to prepare the data for training to effectively save time and costs, and for deployment, both “cloud” and “on-premises” which are unique in the market are available as an option.
HPC → AIHPC Large-scale GPU parallel computing has been used to train the first large-scale successfully enhanced traditional Chinese language model.
A cross-node parallel computing environment is provided for precise model segmentation. InfiniBand architecture and efficient distributed training will deliver near-perfect high-performance validation with cross-node linear performance. With 768 GPUs and the AIHPC supercomputer environment, models are trained successfully with up to 176 billion parameters and a performance close to the “Formosa Foundation Model” of GPT 3.5.
OneAI → No-Code Platform Get started easily with all-in-one AI/MLOps platform tools and a variety of Hugging Face open-source models.
For easy operation, a subscription service for the No-code AI tool platform is provided, with a variety of Transformer-based Models from Hugging Face. This reduces the technical barriers and helps users and teams quickly create and manage the lifecycle of AI solutions.
Build Your Own IndustrialGPT Now
Free Consultation Service
Contact our experts to learn more and get started with the best solution for you.