MINIX, a pioneer in compact high-performance computing, today officially announces the MINIX T4000 and T5000 Generative AI Mini Workstation—powerful, space-saving solutions built for professional generative AI, local large language model (LLM) inference, content creation, on-premise enterprise deployment, and lightweight model training.
Powered by the NVIDIA® Jetson AGX Thor series modules (T4000 / T5000) with the flagship Blackwell architecture, the T4000 / T5000 delivers exceptional on-device AI horsepower in a small desktop footprint. Engineered for professionals, developers, creators, and IT teams, this mini workstation redefines edge and on-premise AI without bulky server hardware.
Next-Gen NVIDIA Blackwell Architecture
At the core of the T4000 / T5000 lies NVIDIA’s cutting-edge compute platform:
- T4000: 1200 Sparse FP4 TFLOPs AI performance
- T5000: Up to 2070 Sparse FP4 TFLOPs AI performance
- 1536–2560 core Blackwell GPU with fifth-generation Tensor Cores
- Multi-Instance GPU (MIG) for parallel task efficiency
- NVIDIA PVA 3.0 dedicated vision processing engine
The workstation natively supports smooth local inference for 7B–70B parameter large language models, making private, low-latency AI accessible for businesses and creators.
Powerful CPU & High-Bandwidth Memory
To eliminate bottlenecks in large model operations, the T4000 / T5000 features high-core-count Arm processing and large, fast memory:
- 12-core (T4000) or 14-core (T5000) Arm® Neoverse®-V3AE 64-bit CPU
- Up to 128 GB high-bandwidth LPDDR5X (4266 MHz, 273 GB/s)
- Massive memory headroom for LLMs, multi-modal AI, and concurrent workloads
High-Speed Connectivity & Rich I/O
Designed for professional workflows, the mini workstation includes enterprise-grade networking and flexible expansion:
- Dual 10GbE Ethernet for ultra-fast data transfer
- Wi-Fi 6E + Bluetooth 5.3 for stable wireless connectivity
- 2× HDMI 2.1 TMDS (4K@60 Hz) for dual high-resolution displays
- 4× USB 3.2 Gen 1 Type-A + 1× USB 3.2 Gen 2 Type-C
- 24V DC input, up to 200W max power
Massive Storage & Compact Design
The T4000 / T5000 balances performance, storage, and space efficiency:
- 1 TB PCIe 4.0 NVMe M.2 SSD (Upgradable up to 4TB)
- Compact dimensions: 3 × 131 × 76.8 mm
- Lightweight: 1420 g
- Durable metal + plastic chassis
- Twin turbo intercooler for sustained high-load performance
- No rack required — plug–and–play desktop deployment
On-Premise AI & Full NVIDIA Software Ecosystem
Security, privacy, and compatibility are central to the T4000 / T5000 design:
- Local private deployment: All data and models processed on-device, no cloud upload needed
- Pre-installed Ubuntu 24.04 LTS (NVIDIA JetPack 7.1)
- Full support for CUDA, TensorRT, NIM, and containerized AI workflows
- Industrial-grade reliability: 255-level programmable watchdog, ESD immunity Grade 4, and wide environmental tolerance
Ideal Use Cases
The MINIX T4000 / T5000 is built for professional generative AI workflows:
- Local LLM Inference: Private, low-latency AI chat, reasoning, and document processing
- Generative AI Creation: AI image, video, 3D, and digital content production
- On-Premise AI Computing: Enterprise private AI hub, secure team AI services
- Lightweight Model Training: LoRA fine-tuning, model distillation, and dataset processing
