Microsoft is all in with its current efforts to expand the use of generative AI in nearly all aspects of its business, from Bing Chat to Copilot for Windows and more. However, the company says it also wants to give startup companies who want to develop generative AI products access to some of the same hardware Microsoft has for its own AI service, and for free.
This week, Microsoft announced plans to offer startups access to its Azure AI services, which include being able to use its GPU virtual machine clusters. Startups can access this hardware in order to train and run large language models (LLMs) for their products.
At first, Microsoft will work with the famed startup funding company Y Combinator to offer these new and free AI services to a limited number of the group"s companies in a private preview. Microsoft"s own venture fund, M12, will also allow its startup companies to get access to the company"s AI hardware. There"s no word yet on when these new startup programs will become generally available.
Microsoft added:
On top of world-class infrastructure, we will also provide tools to simplify deployment and management through Azure Machine Learning. This enables easy low-code or code-based training of custom models and fine-tuning of frontier and open-source models, simplified deployment and optimizations like Low Rank Adaptation, DeepSpeed and ONNX Runtime (ORT). Further, startups can deploy AI solutions with peace of mind knowing all deployments are secure and backed by Microsoft’s principles for Responsible AI.
The company already launched the Microsoft for Startups Founders Hub in 2022. It offers qualified tech startups free access to the GitHub developer service along with Microsoft Cloud services. Some of the other benefits include $2,500 of OpenAI credits for free, along with $150,000 in free Azure credits that AI-themed startups could use to access Azure OpenAI Service.