Compute modules

The Turing Pi 2 Board is designed with four vertical 260-pin SO-DIMM sockets. This design gives you the flexibility to mix and match Compute Modules from different models or vendors in any way you prefer.

For example, you might start with a Kubernetes cluster composed of four Turing RK1 modules. But the beauty of this setup is that you can easily swap out one or more of these RK1 modules with Nvidia Jetson modules. This way, your cluster instantly gains nodes capable of Machine Learning tasks. It's all about flexibility and customization to suit your specific needs.

VendorModel
Turing MachinesRK1
Raspberry PiCM4 (Requires adapter board)
NvidiaJetson Orin NX
NvidiaJetson Orin Nano
NvidiaJetson Nano B01 (A02 version is not supported)
NvidiaJetson Xavier NX
NvidiaJetson TX2 NX