> ## Documentation Index
> Fetch the complete documentation index at: https://docs.rminte.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Introduction
> Welcome to the RM-01 AI Supercomputer Documentation - your comprehensive guide to deploying and managing enterprise AI solutions
## What is RM-01?
The RM-01 is a state-of-the-art portable AI supercomputer designed for private, on-premises deployment of large language models and AI applications. Built for enterprise environments, it delivers powerful AI capabilities while maintaining complete data sovereignty and security.
### Key Capabilities
Up to 2070 TFLOPS (FP4) computing power with 32/64/128GB GPU memory for efficient local AI inference
Dedicated 24GB RAM for AI applications and 8TB SSD for RAG data storage
Hardware-level asymmetrical encryption and secure key management providing fully private, on-premises operation with no data leaving your device
0 cost for deployment, maintenance and a flat learning curve making it the most cost-effective local AI solution for enterprises
## Core Features
### High-Performance Local Inferencing with Flexible Model Support
**Supported Model Sizes and Types:**
* 0.5B to 235B parameters (GPTQ Int4 quantization)
* Mainstream open-source models (DeepSeek, Qwen, Llama)
* Custom enterprise/industry-specific models
* Support for multi-modal models
**Inference Capabilities:**
* Concurrent multi-model support
* Multi-user support
* Hardware-accelerated inference with total throughput of over 2000 TPS (tokens per second)
**Deployment Guarantees:**
* Plug-and-play installation
* Real-time model switching
* Zero-maintenance design
* Enterprise integration ready
### One-stop Solution for Enterprise AI
* 24GB dedicated RAM for AI applications
* Managing and deploying AI applications with Docker
* Asymmetrical hardware-level encryption for safe application delivery
* Embedding model automatically loaded and ready for RAG applications
* Reranker model automatically loaded and ready for RAG applications
* Vector database automatically loaded and ready for RAG applications
* 8TB SSD for data storage
* One USB-C port for connection to any device
* RM-01 will be recognized as a network device
* Use and manage AI applications/models with assigned local IP addresses and APIs
* Could be used as a server with a USB-C to Ethernet adapter
### Enterprise-Grade Security & Privacy
All data processing occurs locally on your device. No information is transmitted to external servers or cloud services.
Built-in hardware-level encryption ensures your data remains secure at all times, even the device is lost or stolen.
Enterprise-grade user authentication and authorization systems protect against unauthorized access.
Meets industry standards for data protection and privacy regulations including GDPR, HIPAA, and SOC 2.
### Lightweight Deployment for Low Enterprise TCO
* One-click deployment within 300 seconds with 0 cost
* No dedicated IT staff required
* Not intrusive to existing IT infrastructure
* Extremely stable and reliable with 0 maintenance required
* Power off and on to recover from any unexpected issues
* No dedicated maintenance staff required
* Flat learning curve for enterprise users with no need to learn complex AI technologies
* Card swapping with one-touch deployment for any AI solutions
## What's Next?
Choose your path based on your role and requirements:
**For Enterprise Users**
Get started with your RM-01 device, including hardware setup, management interface access, and using AI applications.
**For Vendors & Service Providers**
Complete deployment guide for implementing RM-01 in enterprise environments, including model preparation and system integration.
**For Developers**
Technical guide for building applications on RM-01, covering system architecture, environment setup, and model integration.
***