NVIDIA

NVIDIA 940-25840-0011-000 | IGX Thor T7000 Dev Kit, Blackwell GPU, 5581 TFLOPS AI

SKU:940-25840-0011-000

Stock Status: Enquire

Request Quote
Sale Sold out
Shipping calculated at checkout.
20-min response

Need a quote or bulk pricing? Answered within 20 minutes, Sydney business hours.

Description

The NVIDIA IGX Thor T7000 Developer Kit is an enterprise-ready edge AI platform featuring Blackwell architecture iGPU with 2560 CUDA cores, 14-core Arm Neoverse-V3AE CPU, and includes an RTX PRO 6000 Blackwell dGPU for up to 5581 FP4 TFLOPS of AI compute. Built for industrial, medical, and robotics applications requiring real-time sensor processing, functional safety compliance, and advanced AI reasoning at the edge. Includes ConnectX-7 SmartNIC with dual 200GbE connectivity, 128GB LPDDR5X memory, 1TB NVMe storage, BMC, and comprehensive I/O for autonomous systems development.

Features

NVIDIA Blackwell architecture iGPU with 2560 CUDA cores and 96 fifth-generation Tensor Cores
- Pre-installed NVIDIA RTX PRO 6000 Blackwell Max-Q dGPU for 5581 FP4 TFLOPS combined AI performance
- 14-core Arm Neoverse-V3AE CPU for real-time, deterministic processing
- 128GB LPDDR5X memory with 273 GB/s bandwidth for large AI model support
- Multi-Instance GPU technology for concurrent AI workload execution with isolation
- Transformer Engine optimized for generative AI and large language model inference
- Functional Safety Island integrated into Thor SoC for ISO 26262 and IEC 61508 compliance
- Safety MCU and Board Management Controller for mission-critical reliability
- NVIDIA ConnectX-7 SmartNIC with 400GbE networking and RDMA support
- Dual 200GbE QSFP112 ports for high-bandwidth sensor fusion
- Third-generation Programmable Vision Accelerator for real-time computer vision
- Dual hardware video encoders and decoders supporting multiple 4K and 8K streams
- Optical flow accelerator for motion tracking and sensor processing
- PCIe Gen5 expansion support for additional accelerators and storage
- 1TB NVMe storage for model deployment and data logging
- Wi-Fi 6E and Bluetooth for wireless connectivity
- Industrial-grade design with 10-year product lifecycle support
- Compatible with NVIDIA Isaac GR00T, Cosmos, and other physical AI frameworks
- Supports NVIDIA Halos Outside-in Safety Agent Blueprint for industrial safety applications
- MicroATX form factor for flexible edge deployment configurations

Warranty

All products sold by XS Network Tech include a 12-month warranty on both new and used items. Our in-house technical team thoroughly tests used hardware prior to sale to ensure enterprise-grade reliability.

All technical data should be verified on the manufacturer data sheets.

View full details

specs-tabs

Collapsible content

Technical Specifications

FAQs

Technical Specifications

GPU: 2560-core NVIDIA Blackwell architecture iGPU with 96 fifth-generation Tensor Cores, Multi-Instance GPU support
- Discrete GPU: NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition pre-installed
- AI Performance: Up to 5581 FP4 TFLOPS combined (2070 TFLOPS iGPU + 3511 TFLOPS dGPU)
- CPU: 14-core Arm Neoverse-V3AE 64-bit processor
- Memory: 128GB 256-bit LPDDR5X, 273 GB/s bandwidth
- Storage: 1TB NVMe SSD (M.2 Key-M slot)
- Networking: NVIDIA ConnectX-7 SmartNIC with dual QSFP112 ports (2x 200GbE), 400GbE total bandwidth, PCIe Gen5 32-lane switch
- Additional Networking: 2x 1GbE RJ45 ports, Wi-Fi 6E (802.11ax), Bluetooth
- I/O Ports: 4x USB-A 3.2, USB-C 3.1, DisplayPort 1.4a, HDMI, CAN headers
- Form Factor: MicroATX with uATX power supply, desktop/edge deployment chassis
- Power: 40W-130W TMP (Thor Module Package), up to 300W for dGPU
- Safety Features: Functional Safety Island on SoC, Safety MCU, Board Management Controller
- Accelerators: Third-generation Programmable Vision Accelerator, dual video encoders/decoders, optical flow accelerator
- Functional Safety: Designed for ISO 26262 ASIL D and IEC 61508 SIL 3 compliance
- Software Stack: NVIDIA AI Enterprise compatible, supports Isaac, Metropolis, Holoscan, NIM frameworks

FAQs

Q: What are the primary use cases for the IGX Thor T7000 Developer Kit?
A: The IGX Thor T7000 Developer Kit is designed for industrial edge AI, medical imaging, surgical robotics, humanoid robotics, collaborative robots, autonomous machines, and high-performance edge computing applications requiring functional safety. It excels at generative AI reasoning, multimodal sensor fusion, real-time vision processing, and running multiple AI models simultaneously with workload isolation via Multi-Instance GPU support.

Q: How does the IGX Thor T7000 compare to the previous IGX Orin platform?
A: The IGX Thor T7000 delivers up to 8x higher AI compute on the integrated GPU, 2.5x higher AI compute with discrete GPU, and 2x better connectivity compared to IGX Orin. The Blackwell architecture provides 5581 FP4 TFLOPS versus Orin's 1705 INT8 TOPS, enabling significantly faster generative AI inference and support for larger language and vision models at the edge.

Q: What networking capabilities does the ConnectX-7 SmartNIC provide?
A: The integrated NVIDIA ConnectX-7 SmartNIC delivers 400GbE total bandwidth via dual 200GbE QSFP112 ports, providing 2x the bandwidth of previous generations. It supports RDMA for low-latency, high-throughput data movement essential for sensor-intensive workloads like LIDAR, multiple camera arrays, and real-time medical imaging systems. The 32-lane PCIe Gen5 switch enables direct high-speed connectivity to multiple sensors and devices.

Q: Can the IGX Thor T7000 run multiple AI models simultaneously?
A: Yes, the IGX Thor T7000 leverages Blackwell architecture features including Transformer Engine and Multi-Instance GPU technology to run multiple generative AI models concurrently with robust workload isolation. This enables complex applications that require simultaneous vision-language-action models, LLMs, VLMs, and real-time inference while maintaining deterministic performance and functional safety requirements.

Q: What software frameworks and tools are supported?
A: The IGX Thor T7000 runs the NVIDIA AI Enterprise software stack and supports NVIDIA Isaac for robotics, Metropolis for visual AI agents, Holoscan for sensor processing, and NIM for physical AI deployment. It provides a complete cloud-to-edge development experience with containerized workflows, allowing developers to build on the IGX Thor Developer Kit and deploy to production T7000 systems without code changes.

Recently Viewed

  • Request a Quote

    Looking for competitive pricing? Submit a request, and our team will provide a tailored quote that fits your needs.

  • Contact Us Directly

    Have a question or need immediate assistance? Call us for expert advice and real-time support.

    Call us Now  
  • Contact Us Directly

    Have a question or need immediate assistance? Call us for expert advice and real-time support.

    Contact us