Apple’s presence at AWS re:Invent 2024 surprised everyone, as the company has traditionally been known for its closed ecosystem. During his keynote, AWS chief Matt Garman invited Benoit Dupin, Apple’s senior director of machine learning and AI, on stage to speak about the company’s partnership with AWS and how it has helped power its AI and machine learning features.
Dupin highlighted the importance of the decade-long partnership with AWS in scaling Apple’s ML and AI capabilities. He oversees machine learning, AI, and search infrastructure at Apple and explained how the company’s AI-driven features, such as Siri, iCloud Music, and Apple TV, heavily rely on AWS’s infrastructure.
According to Dupin, AWS has consistently supported Apple’s dynamic needs at scale and globally. The company has increasingly leveraged AWS’s solutions, including its Graviton and Inferentia chips, to boost efficiency and performance. Dupin revealed that Apple achieved a 40% efficiency gain by migrating from x86 to Graviton instances. Additionally, transitioning to Inferentia 2 for specific search-related tasks enabled the company to execute features twice as efficiently.
This year, Apple launched Apple Intelligence, which integrates AI-driven features across iPhone, iPad, and Mac. Dupin explained that Apple Intelligence is powered by the company’s own large language models, diffusion models, and adapts on both devices and servers. Key features include system-wide writing tools, notification summaries, and improvements to Siri, all developed with a focus on user privacy. To support this innovation, Apple required scalable infrastructure for model training and deployment.
Dupin acknowledged that AWS services have been instrumental in supporting Apple’s AI and ML lifecycle, including fine-tuning models and building adapters for deployment. The company is also exploring AWS’s Trainium2 chips, with early evaluations suggesting a 50% improvement in pre-training efficiency. Dupin emphasized that AWS’s expertise, guidance, and services have been critical in supporting Apple’s scale and growth.
In a technical paper published by Apple on Monday, the company revealed that it uses Google’s Tensor Processing Units (TPUs) instead of the industry-standard NVIDIA GPUs for training its AI models. However, at AWS re:Invent 2024, Amazon Web Services (AWS) announced the general availability of AWS Trainium2-powered Amazon Elastic Compute Cloud (EC2) instances. These new instances offer 30-40% better price performance than the previous generation of GPU-based EC2 instances.
In conclusion, Apple’s partnership with AWS has been crucial in scaling its AI and ML capabilities, and the company’s recent launch of Apple Intelligence is a testament to this. With the help of AWS’s solutions and expertise, Apple has been able to achieve significant efficiency gains and develop innovative features while prioritizing user privacy. The availability of AWS Trainium2-powered EC2 instances further strengthens this partnership and highlights the importance of scalable infrastructure in driving AI and ML advancements.
