Olir AI
-
Introduction
Olir AI provides automakers additional revenue from underutilized in-car GPUs while offering AI companies lower-cost solutions for asynchronous inferencing tasks.
Problem
Car manufacturers face the challenge of marginal profits from car sales and are seeking ways to boost lifecycle revenue through increased utilization of vehicles and components. Simultaneously, the rapid rise of large language models (LLMs) has created an enormous demand for GPUs to support AI training and inferencing. This surge has significantly driven up costs for startups and companies needing access to high-performance GPUs, making it harder to scale AI-driven solutions. Our platform addresses both needs by optimizing GPU benchmarking for AI processes, reducing costs, and enabling efficient resource allocation, thus empowering car manufacturers to enhance profitability while driving innovation.
Opportunity
We propose creating a marketplace to leverage the unused compute power of Nvidia GPUs in electric vehicles (EVs) and rent it to AI companies. Many automakers have partnered with Nvidia to equip their cars with Drive Orin GPUs, designed for autonomous driving. These GPUs excel at video and image-based AI training and inferencing, making them ideal for asynchronous inferencing tasks. By tapping into this underutilized resource, we can offer AI companies the same performance as current GPU providers at a significantly lower cost, while helping automakers generate additional revenue from their vehicles.