A detailed walkthrough of building a budget-friendly AI workstation with 48GB VRAM for running local LLMs, costing around 1700 euros using second-hand Tesla P40 GPUs. The setup enables running various AI models locally, achieving 5-15 tokens per second depending on model size, while maintaining independence from cloud-based AI services.
An exploration of 'frupidity' - the harmful practice of misguided cost-cutting that ultimately damages productivity and innovation in engineering organizations. The article examines how excessive frugality, coupled with bureaucracy, leads to hidden costs through reduced productivity, lowered morale, and lost opportunities, while providing practical strategies to combat this organizational pattern.