On October 17, it was announced that Arm, a global semiconductor intellectual property (IP) company, has strengthened its strategic partnership with Meta. This collaboration aims to expand AI efficiency across the entire computing spectrum, from AI software to data center infrastructure. The partnership intends to provide an enhanced AI experience to billions of users worldwide by optimizing AI performance across computing environments that power Meta’s global platform, ranging from milliwatt-level low-power devices supporting on-device intelligence to megawatt-level large-scale systems responsible for training advanced AI models.
This strategic partnership builds on years of joint design collaboration in hardware and software. By combining Arm’s power-efficient AI computing technology with Meta’s AI products, infrastructure, and open-source technology innovation, the goal is to significantly enhance performance and efficiency.
Meta plans to utilize Arm Neoverse-based data center platforms to support AI-powered ranking and recommendation systems across its app portfolio, including Facebook and Instagram. This is expected to provide high performance and low power consumption compared to x86 systems, thereby achieving power-performance efficiency throughout Meta’s infrastructure and enhancing the efficiency and scalability of Arm computing in hyperscale environments.
Moreover, this partnership aims to further optimize AI software across the PyTorch machine learning framework, ExecuTorch edge inference runtime engine, and vLLM data center inference engine. The optimized ExecuTorch base by Arm KlediAi will enhance efficiency on billions of devices and simplify model deployment, accelerating AI application performance from edge to cloud.
The full-stack collaboration between Arm and Meta spans the computing spectrum from megawatt-level data centers to core AI software, expanding AI capabilities. This partnership is set to deliver a more intelligent, efficient, and connected experience to billions worldwide.
René Haas, CEO of Arm, stated, “The next-generation AI era will focus on achieving large-scale efficiency. Through our collaboration with Meta, we will combine Arm’s performance-per-watt leadership with Meta’s AI innovation capabilities to realize smarter and more efficient intelligence in every domain, from milliwatt to megawatt levels.”