OpenAI plans to purchase AI chips worth $51 million from a startup supported by CEO Sam Altman
December 3 2023
Sam Altman, previously fired as the CEO of OpenAI, was involved in a deal where OpenAI intended to purchase $51 million worth of AI chips from Rain AI, a company he personally invested in. The deal illustrates potential conflicts between Altman’s personal investments and his role at OpenAI. Rain, which works on neuromorphic processing units (NPUs), focuses on advanced chip technologies that could greatly improve computing power and energy efficiency for AI development compared to traditional GPUs. OpenAI’s commitment to this purchase underscores its need for cutting-edge hardware to support AI advancements, despite existing hardware constraints and reliance on major cloud providers like Microsoft. Altman has also explored establishing a new AI chip company to address global supply issues. However, Rain’s progress faced scrutiny after a Saudi-affiliated fund’s investment was blocked by the US government over national security concerns, which could delay OpenAI’s access to these chips. Additionally, Rain’s leadership has seen changes, with CEO Gordon Wilson stepping down. Despite these challenges, OpenAI’s need for innovative hardware solutions remains critical as the company looks to maintain a competitive edge in AI technology development.
Back to Breaking AI News
What does it mean?
- CEO: The chief executive officer, the highest-ranking person in a company or other institution, ultimately responsible for making managerial decisions.
- OpenAI: A research institute that specializes in developing artificial intelligence in a safe and beneficial manner.
- AI chips: Specialized processors designed specifically to facilitate and optimize the computations required to run artificial intelligence algorithms more efficiently than general-purpose CPUs.
- Neuromorphic processing units (NPUs): A type of computer technology that is designed to mimic neural networks of the human brain, potentially allowing for more efficient processing in tasks related to artificial intelligence and machine learning.
- GPUs: Graphics processing units, which are processors originally designed to render graphics but are also used for mathematical calculations required in AI computations due to their parallel processing capabilities.
- Hardware constraints: Limitations in the physical components of computing infrastructure, which can hamper performance and scalability of technological applications.
- Cloud providers: Companies that offer cloud computing services, which include servers, storage, databases, networking, software, analytics, and intelligence, over the internet ("the cloud") to offer faster innovation, flexible resources, and economies of scale.