Nvidia quietly unveiled its fastest mini PC ever, capable of topping 2070 TFLOPS – and if you squint enough, you might even think it looks like an RTX 5090
- Jetson T5000 claims 2070 TFLOPS performance through Blackwell GPU architecture
- Jetson AGX Thor Developer Kit transforms a compact board into workstation-level computing
- T4000 is positioned as a lighter, cost-efficient alternative option
Nvidia has expanded its Jetson lineup with the Jetson AGX Thor Developer Kit, a compact platform that carries the new Jetson T5000 system-on-module.
Marketed as a developer system, the dimensions and form factor place it firmly in the realm of a mini PC, although its design and purpose align more with edge AI deployment than home computing.
Nvidia says the Jetson T5000 delivers “2070 TFLOPS (FP4, Sparse),” made possible by its 2560-core GPU based on the Blackwell architecture, with 96 fifth-generation Tensor Cores and Multi-Instance GPU features.
Raw power behind the Jetson T5000
This system is paired with a 14-core Arm Neoverse-V3AE CPU and 128GB of LPDDR5X memory.
Networking is handled by four 25GbE connections, with support for NVMe storage through PCIe.
The Jetson AGX Thor kit includes video encode and decode support across multiple 4K and 8K streams.
There is also a lower-end option, the Jetson T4000, which is still in development, but early specifications list “1200 TFLOPS (FP4, Sparse)” performance, a 1536-core GPU, and 64GB of memory.
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
Both modules operate across a wide power range, with the T5000 rated between 40 and 130 watts and the T4000 between 40 and 75 watts.
This device is designed to provide researchers and engineers with a complete platform for testing robotics and edge workloads.
For connectivity, it ships with a reference carrier board equipped with a WiFi 6E module, 1TB NVMe SSD, and standard debugging interfaces.
Networking includes a QSFP28 interface with four 25GbE channels and a 5GbE RJ45 connector, highlighting its focus on sensor-heavy applications.
The kit also supports expansion through M.2 slots and offers HDMI 2.0b and DisplayPort 1.4a outputs, along with multiple USB ports.
Its physical dimensions are 243.19 x 112.4 x 56.88 mm, making it larger than a business PC but still compact compared with most workstation PC designs.
Nvidia positions this release alongside earlier initiatives such as the DGX Spark, which was presented as a desktop AI development platform.
The Jetson AGX Thor differs by targeting humanoid robotics, visual AI systems, and sensor integration, supported by the company’s Isaac, Metropolis, and Holoscan software frameworks.
The Jetson AGX Thor Developer Kit is listed at $3,499 and is available for pre-order from selected distributors, with shipments expected to begin on November 20, 2025.
Via Cnx-software
You might also like
Jetson T5000 claims 2070 TFLOPS performance through Blackwell GPU architecture Jetson AGX Thor Developer Kit transforms a compact board into workstation-level computing T4000 is positioned as a lighter, cost-efficient alternative option Nvidia has expanded its Jetson lineup with the Jetson AGX Thor Developer Kit, a compact platform that carries the…
Recent Posts
- Steam Machine and Steam Frame are coming ‘this summer’
- Valve says it’s ready to launch the Steam Machine this summer
- Best Buy slashes up to $400 off Apple tech in a limited-time sale — get AirPods, MacBooks, iPads and Apple Watches from $99.99
- The Instagram Plus subscription has officially launched
- Wired found code for an unreleased facial recognition feature in Meta’s AI app
Archives
- June 2026
- May 2026
- April 2026
- March 2026
- February 2026
- January 2026
- December 2025
- November 2025
- October 2025
- September 2025
- August 2025
- July 2025
- June 2025
- May 2025
- April 2025
- March 2025
- February 2025
- January 2025
- December 2024
- November 2024
- October 2024
- September 2024
- August 2024
- July 2024
- June 2024
- May 2024
- April 2024
- March 2024
- February 2024
- January 2024
- December 2023
- November 2023
- October 2023
- September 2023
- August 2023
- July 2023
- June 2023