Tesla(US:TSLA) Zhong Guo Qi Che Bao Wang·2026-01-23 04:30Core Viewpoint - Tesla has announced the restart of the Dojo 3 supercomputer project, which is crucial for the development of AI chips and autonomous driving technology [2][6]. Group 1: Importance of Supercomputers - Supercomputers are referred to as the "Everest" of computing, designed for high-speed calculations, large storage capacity, and high communication bandwidth, making them essential for processing large datasets in various applications [3]. - The demand for computing power in autonomous driving is described as "massive," with key metrics like 1 EFLOPS and PB-level storage becoming benchmarks for performance [4]. - Supercomputers play a critical role in training neural networks for autonomous driving, enabling the processing of vast amounts of data and complex scenarios [6]. Group 2: Role in Autonomous Driving - The processing capabilities of supercomputers are vital for the development of autonomous driving systems, allowing for real-time data analysis and decision-making [6][7]. - Supercomputers are seen as accelerators of technological iteration in autonomous driving, significantly impacting the commercialization of autonomous taxi projects [7]. - The core value of supercomputers in autonomous driving lies in their ability to overcome key bottlenecks in data processing, model training, and algorithm optimization [8]. Group 3: Challenges and Costs - Despite their advantages, supercomputers come with high construction and operational costs, potentially reaching hundreds of millions of dollars, which can be a burden for some automotive companies [8]. - The successful implementation of autonomous driving technology requires not only powerful computing resources but also supportive regulations, infrastructure, and data security measures [8]. Group 4: Competitive Landscape - A global competition for supercomputing power is underway, positioning supercomputers as a core competitive barrier in the autonomous driving sector [9]. - Companies with robust supercomputing capabilities can gain a competitive edge in the development and optimization of autonomous driving technologies [9].