Overview
The NVIDIA IGX T7000 is the production-grade variant of the IGX Thor platform, designed for deployment in industrial and medical environments. It delivers the same 5,581 TFLOPS of AI performance as the developer kit, packaged in a MicroATX board form factor with BMC (Baseboard Management Controller) for enterprise-grade remote management.
Key Features
- 5,581 TFLOPS FP4 Sparse: Full IGX Thor AI performance in production form factor
- RTX PRO 6000 dGPU: 24,064 CUDA cores for maximum edge AI compute
- BMC: Enterprise remote management and monitoring
- MicroATX Form Factor: Standard board for integration into industrial enclosures
- ConnectX-7: 200GbE networking for high-throughput data pipelines
- 4x SATA + NVMe: Flexible storage for production data logging
Technical Specifications
| Specification | Details |
|---|---|
| iGPU | 2,560-core NVIDIA Blackwell |
| dGPU | NVIDIA RTX PRO 6000 (24,064 CUDA cores) |
| CPU | 14-core Arm Neoverse-V3AE @ 2.6 GHz |
| System Memory | 128 GB LPDDR5X |
| GPU Memory | 96 GB GDDR7 |
| AI Performance | 5,581 TFLOPS (FP4 Sparse) |
| Storage | 4x SATA + M.2 NVMe |
| Networking | ConnectX-7, 2x 1GbE, 2x 200GbE |
| Management | BMC with dedicated port |
| Safety | ISO 26262 ASIL D / IEC 61508 SIL 2 |
| Form Factor | MicroATX (244 x 199 x 31 mm) |
Ideal Use Cases
- Production medical AI systems — surgical navigation, diagnostic imaging, robotic-assisted procedures
- Industrial automation controllers with functional safety requirements
- Fleet-deployed AI inference servers at the industrial edge
- Critical infrastructure monitoring with enterprise manageability
Why Choose This Product?
The IGX T7000 is built for production. With BMC for remote management, MicroATX compatibility for standard enclosures, and the full safety certification of the IGX Thor platform, it is the system you deploy when your industrial AI application moves from prototype to production.
Interested? Contact us for production quantities, custom enclosure design, and deployment support.
