On-Premises AI Solution
The optimal generative AI solution for enterprises—ready to deploy with secure and fully controllable operations.
Popular Searches
AI² × aiDAPTIV+
One-Stop On-Premise AI Solution
Taiwan Mobile partners with Phison to integrate software and hardware strengths, delivering a one-stop AI solution for development, training, and deployment. This enables enterprises to quickly adopt generative AI with easy setup, private and secure operation, and optimized performance
aiDAPTIV+ Solution
Scalable Model Computing Performance
Lower the Barrier to Enterprise AI Adoption
aiDAPTIVCache overcomes memory limits by using NAND storage as extended memory. Supports larger-scale large language model training. Significantly improves token processing and long-content generation/understanding. Achieves local training and inference with fewer GPUs and lower costs.
An AI-powered transcription tool that understands local languages.
Accurately recognizes Mandarin, English, Taiwanese, and Hakka in seconds, automatically generating transcripts and summaries so key information is captured instantly.
See more details
查看說明
Enterprise-grade localized speech recognition models.
Developed in collaboration with government agencies and research institutions, trained on millions of Mandarin, Taiwanese, and Hakka language datasets.
See more details
查看說明
Enhance enterprise knowledge with intelligent AI responses.
Powered by LLMs for advanced semantic understanding and integrated with enterprise knowledge bases for fast retrieval and more accurate conversational services.
See more details
查看說明
Simple to Use, Quick to Learn
- No need for extensive IT staff deployment and maintenance
- Ready to use out of the box with graphical interface for team focus on AI applications and innovation
High Performance, High Value
- Uses cost-effective NAND flash memory technology
- Supports LLM training on standard workstation hardware to reduce setup costs and barriers
Local Deployment, Secure Data
- All data stays internal to meet security and compliance requirements
- Full control over models and data assets to prevent information leaks
AI Meeting Records
Automatically records and summarizes meeting content to improve efficiency and information retention.
AI Voice Customer Service
Handles interactions via voice to shorten processing time and create tailored dialogue experiences.
AI Chat Customer Service
Chatbots respond quickly to user questions, reducing staff workload and increasing satisfaction.
AI Assistant
Super assistant with task execution capabilities for handling and querying tasks to boost work efficiency.
Powered by GPUs and Phison aiDAPTIV+ technology, enabling large language model (LLM) training and efficient local inference
Enterprise AI Server
1.Supports training for models with up to 405B parameters
2.No strict limitations on model size
One-stop environment for large-scale model training
3.Flexible node expansion with near-linear performance scaling
2.No strict limitations on model size
One-stop environment for large-scale model training
3.Flexible node expansion with near-linear performance scaling
High-Performance AI Workstation
1.Supports training for models with up to 100B parameters
2.Lightweight deployment with rapid setup
3.Ideal for departmental use and pilot deployments
4.Enables secure on-premises data processing
2.Lightweight deployment with rapid setup
3.Ideal for departmental use and pilot deployments
4.Enables secure on-premises data processing
