The EU has opened an Antitrust Investigation into Corning over Exclusive Clauses regarding contracts for Device Cover Glass
In Q3 2024 Three Apple iPhone 15 models were the best selling Globally and more

Foxconn is only able to produce a limited amount of AI servers for Apple in Taiwan due to Nvidia’s domination of Foxconn’s capacity

1 Apple Intelligence

In order to speed up the process of gaining a larger market share in the generative AI boom, Apple is rumored to have approached Hon Hai (Foxconn) with the hope of producing AI servers in Taiwan, according to a new Nikkei Asia report.

Hon Hai is Apple's largest iPhone manufacturing partner, but it is also a manufacturer of Nvidia AI servers. Nikkei quoted sources as saying that Hon Hai's ability to take orders for Apple's AI servers is limited.

Apple is keen to use its own chips to build servers that support generative AI features in MacBooks and other Apple devices. Sources pointed out that since Apple hopes to produce its own servers, the production quantity discussed with Hon Hai is relatively low compared to the demand for Nvidia GB200 servers.

The report also quoted a person with direct knowledge of the situation as saying that one of the reasons why Apple wants Hon Hai to produce AI servers in Taiwan is to use engineering talents and R&D resources suitable for the project. Apple has less experience designing data center servers than Nvidia, so it may need more support from vendors to provide engineering design services for the servers.

With today's technological trends, mastering AI computing and expanding data center infrastructure has become the key to next-generation computing. Apple is beefing up its server capabilities as rivals such as Microsoft, Amazon and Google invest heavily in artificial intelligence data centers.

According to people familiar with the matter, as Hon Hai is busy completing orders for Nvidia servers, Apple has also asked China's Lenovo and its subsidiary LCFC to provide partial assistance, and hopes that smaller suppliers such as USI Electronics will also provide production services. The negotiations between Apple and Lenovo also involve the creation of new server production capacity in Southeast Asia outside China.

The Nikkei report further quoted industry insiders as saying that manufacturing AI servers has proven to be more complicated than expected because the process requires a complete overhaul of key components and production line systems. The integration of new components (such as cooling technology) into hardware and software is also more complex than in traditional servers. It is said that the cooling system of just one server production line costs tens of millions of dollars.

Currently, only a handful of certified server manufacturers in the world have the ability to produce Nvidia's AI servers, including Hon Hai, Quanta, Wistron, and Supermicro, and most of the production capacity of these manufacturers is now being used by Nvidia. All inclusive.

Apple’s Apple Intelligence function is one of the key selling points of this year’s new iPhone 16 series. Its generative AI functions include helping users retouch the content of documents/letters, customize emoticons, and create images by entering some descriptive text. Siri has also been upgraded to allow users to use more natural language when issuing commands.

The report pointed out that the appropriate negotiations between Hon Hai and Apple on the manufacturing of AI servers have not yet been finalized, because Hon Hai’s current AI server production capacity in Hsinchu is all contracted by Nvidia GB200; and in addition to the space and equipment required to fulfill orders, the space and equipment required to fulfill the order are limited. In addition, the two parties have not yet reached an agreement on the price.

10.0F0  Supply Chain News & Rumors