Apple is reportedly set to purchase thousands of high-end AI servers over the next two years and massively boost infrastructure spending to train generative AI models, according to noted Apple analyst.
Apple to Buy Up to 20,000 Nvidia Servers
Apple could acquire between 2,000-3,000 Nvidia HGX AI supercomputers this year, rising to as many as 20,000 servers in 2023. Each server costs around $250,000.
Apple is acquiring the powerful servers, equipped with advanced GPUs, to train large language models in-house with maximum security.
Seeks to Bolster Privacy for AI Workloads
By handling intensive AI workloads like generative AI model training internally, Apple aims to improve privacy protections compared to relying on public cloud services.
Bringing more capabilities in-house allows Apple to better control access to potentially sensitive training data. It aligns with Apple’s privacy-focused brand.
But Lags Behind Big Tech Rivals
However, even with the ramped up investments, Apple’s AI infrastructure will significantly trail leading tech giants like Meta and Microsoft who operate massive data centers. Apple risks falling further behind without spending several billion dollars per year on servers, labor, facilities and other AI costs. Competition in AI is fierce.
Could Develop Own AI Chips to Cut Costs
Looking ahead, Apple may design custom chips optimized for AI workloads to reduce server expenses, according to analyst.
currently sees no evidence Apple is yet developing proprietary AI accelerator chips. For now, it appears content using Nvidia’s proven technology.
Overall Spending Estimates Vary
Estimates for Apple’s total AI spending vary. analyst thinks billions per year will be required to be competitive.
But Bloomberg’s Mark Gurman expects a more conservative $1 billion annual investment. Gurman reports Apple aims to integrate AI across its products and improve Siri.
AI Critical to Apple’s Future
How much Apple spends on AI has major implications. AI powers key features like Siri, Photos, facial recognition and more.
Falling behind in generative AI could harm Apple’s ability to match competitors rolling out new AI-driven services. Its future success hinges on AI innovation.
In summary, Apple is reportedly making big moves into AI hardware to train next-gen models locally rather than on the cloud. While costly, the investment is critical as AI becomes integral to staying competitive.
