Mini PCs and Industrial PCs for Machine Learning

Choosing a PC for Machine Learning

A PC for machine learning needs enough processor cores, memory and storage to load datasets and run training code. PyTorch and TensorFlow are the two frameworks named in the facts for this query. The facts also state that PyTorch workloads are GPU-bound and that NVIDIA CUDA, cuDNN and NCCL are its primary acceleration stack. Work that depends on those GPU libraries needs a discrete NVIDIA GPU, which the Thinvent range does not include.

Key Specifications to Check

Machine learning work depends on several specifications:

• Cores and clock speed: More cores run data preparation and CPU-side tasks faster.

• Memory (RAM): Datasets and model inputs are held in RAM during work. 16 GB or more suits most small and medium tasks.

• Storage: Fast SSD storage reduces the time to load datasets.

• Operating system: Windows 11 and Ubuntu Linux are both available on Thinvent products.

Thinvent Range Comparison

Thinvent Products for Machine Learning Workloads

Thinvent has several PCs with the processor cores and memory that suit machine learning development. The Thinvent Industrial PC IPC5 with Intel Core i7-1255U and 32 GB DDR4 RAM has 10 cores, a 4.7 GHz maximum frequency and 1 TB SSD storage. The Thinvent Industrial PC IPC5 with Intel Core i5-1240P has 12 cores and 32 GB RAM. The Thinvent Aero Mini PC with Intel Core 5 120U has 10 cores, 5.0 GHz maximum frequency and 16 GB RAM. The Thinvent Treo Mini PC with AMD Ryzen 5 7430U has 6 cores and 16 GB RAM. For GPU-accelerated PyTorch or TensorFlow training, pair a CPU system with an external compute resource that has an NVIDIA GPU.