Foxconn set to benefit from Nvidia's new venture into AI factories

Ninelu Tu, Taipei; Jack Wu, DIGITIMES Asia 0


In addition to its strong financial results, Nvidia is also actively exploring new business models such as AI factories or AI manufacturing to enhance profitability further. Once Nvidia's AI factory is established, major partners like Foxconn will naturally be one of the first to benefit from it.

Previously, during the Hon Hai Tech Day (HHTD), Nvidia CEO Jensen Huang and Foxconn Chairman Young Liu jointly announced that the two sides would deepen their collaboration in terms of AI factories. Foxconn and Nvidia have been allies since the HGX-1 era, and Foxconn's presence in the AI server domain covers GPU modules, boards, complete units, and racks. Therefore, it's not a surprise that the two companies have decided to team up to promote AI factories.

Regarding the AI waves sweeping through industries, Huang believes that from now on, more and more corporations will want to establish their AI, but they might not be able to afford the costs. Nvidia can leverage its capabilities and technology to help businesses build their AI models and continuously optimize them through AI factories, thus enhancing operational efficiency.

According to Huang's explanation, the AI factory will use Nvidia's computing infrastructure as the core and focus on processing, refining, and transforming large amounts of data into valuable AI models and producing AI solutions. After the AI models are operational, the AI factory will continue to update them based on feedback data from the front lines, making them smarter and more applicable.

Young Liu pointed out that EVs have numerous sensors, and the data collected by these sensors are transformed into tokens. Processing these tokens in the AI factory can produce better software, models, or solutions to make EVs smarter. Besides EVs, Foxconn plans to apply the concept of AI factories to smart cities and smart manufacturing as well.

Market sources pointed out that, whether it's AI factories or AI manufacturing, with the industry transitioning from general-purpose computing to accelerated computing and generative AI, Nvidia is actively developing its new business models. This is especially relevant as more corporations or regions seek to build their AI models and application services, which led to the rise of the AI factory and AI manufacturing concepts.

Huang also mentioned that Nvidia currently provides Nvidia AI Enterprise services on the foundation of the DGX Cloud. Currently, these services may be hosted on the internet architecture of large cloud service providers (CSP). However as the market grows and demand from different businesses arises, these services might become more localized and smaller in the future. It will no longer have to rely on the network of large CSPs, instead being operated by more local providers to cater to the AI demands of regional or small and medium-sized enterprises.

When that time arrives, whoever can provide a more comprehensive hardware platform to help local businesses quickly establish an AI factory architecture and engage in AI manufacturing services will have access to the greatest business opportunities.

Furthermore, Nvidia's developments in the automotive sector are also progressing smoothly. Driven by the autonomous driving platform and AI cockpit solutions, its revenue has continued to grow. The collaboration with Foxconn in the automotive field is also deepening by the day.

Foxconn has previously received authorization to be an Nvidia automotive Tier 1 supplier. Foxconn will utilize the Nvidia DRIVE Orin technology to produce electronic control units (ECUs) for the global automotive market. When the next-generation product DRIVE Thor is launched, Foxconn will also produce ECUs equipped with DRIVE Thor.

As of now, 25 global automakers are using DRIVE Orin. Market sources believe that as demand for automation and self-driving functions increases, the related market is expected to grow, benefiting supply chain members.