Skip to main content
Fig. 8 | Acta Neuropathologica Communications

Fig. 8

From: EpiDiP/NanoDiP: a versatile unsupervised machine learning edge computing platform for epigenomic tumour diagnostics

Fig. 8

Cost- and space-effective, energy-efficient and portable edge computers for NanoDiP. A: Nvidia ORIN CPU/GPU hybrid SoC 32GB developer kit (1) with USB3-attached Mk1B (2) and P2 Solo (3) sequencers. B: Minimalistic cryptocurrency mining mainboard (1) with attached Mk1B sequencer (2) and multiple PCIe connectors in robust USB3 plug format (3) to attach one or more gpGPUs (4) and NVMe modules through riser boards (5) for up-scaling. CPU, RAM, NVMe, and power supply were removed to improve visibility. Components are designed to be mounted on open frames. Scale bar: 5 cm

Back to article page