Major server purchase: Apple about to place AI mega-order with Nvidia?
AI training is not possible without fast hardware. For Apple Intelligence, Apple is now to help itself to the Nvidia shelf once again.
Nvidia logo on a building.
(Image: Michael Vi/Shutterstock.com)
Apple is apparently planning to invest billions in more and newer AI technology for Apple Intelligence. The supplier is said to be GPU specialist Nvidia. This was reported by analyst Ananda Baruah from the American investment firm Loop Capital, citing sources in the supply chain. According to the sources, the order volume is around one billion US dollars. Of this, up to 250 new servers are to be purchased.
AI servers specialized in reasoning
As Baruah writes in an investment report in the stock market journal IBD, the systems in question are of the GB300 NVL72 type. These are systems that are not yet officially on the market and are intended to be particularly suitable for reasoning AI applications. A large language model reveals the work steps that were used to reach a conclusion. Corresponding models are available from major providers such as OpenAI in the USA and DeepSeek in China. Apple does not yet operate its chatbot and does not yet use reasoning in products accessible to customers.
A GB300-NVL72 system is based on 72 Blackwell Ultra GPUs plus 36 Grace CPUs "in a single platform optimized for test-time scaling inference", as Nvidia describes. The systems can also be used to build so-called AI factories, in which the machines are connected to Quantum X800 InfiniBand or Spectrum X Ethernet. When using a ConnectX-8 SuperNIC, the inference output of reasoning models can be increased fifty-fold compared to the previous Hopper platform.
Videos by heise
Large server cluster for Gen AI at Apple
Baruah also announced that Apple plans to spend between 3.7 and 4 million dollars per server. In addition to Nvidia, Dell and Super Micro are probably also among Apple's suppliers for AI hardware. According to the analyst, Apple is now "officially in the game for large server clusters in the Gen AI sector". "Gen AI" stands for generative AI systems, i.e. large language models (LLM) with and without reasoning and image generators.
Apple is pursuing a dual strategy. The company is also developing its own AI servers based on the M4, among others, to implement its private cloud compute concept, in which AI applications run securely in the cloud without the company itself being able to see into them.
Empfohlener redaktioneller Inhalt
Mit Ihrer Zustimmung wird hier ein externer Preisvergleich (heise Preisvergleich) geladen.
Ich bin damit einverstanden, dass mir externe Inhalte angezeigt werden. Damit können personenbezogene Daten an Drittplattformen (heise Preisvergleich) übermittelt werden. Mehr dazu in unserer Datenschutzerklärung.
(bsc)
