OpenAI seeking cloud alternatives

Reportedly, delays in the Stargate project, a joint initiative with Microsoft to build a supercomputer dedicated to artificial intelligence, are pushing OpenAI to explore alternative solutions to meet the growing demand for computational resources.

The company, known for its large language models (LLMs), needs significant computing power to train and perform inference on its models. Delays in the Stargate project could compromise OpenAI's ability to maintain the pace of development and release of new products and services.

The search for cloud alternatives suggests a strategy of diversifying sources of supply for computational resources, potentially including collaborations with other cloud providers or expanding its own infrastructure. For those evaluating on-premise deployments, there are trade-offs that AI-RADAR analyzes in detail at /llm-onpremise.

General context on the importance of computing for AI

Artificial intelligence, and in particular deep learning models, require an enormous amount of computing power. Training complex models like GPT-4 requires clusters of high-end GPUs and specialized infrastructure. The availability and cost of these resources are critical factors for companies operating in the AI ​​sector.