A supercomputer called ChatGPT cost Microsoft hundreds of millions of dollars to build
According to a Bloomberg article, Microsoft invested hundreds of millions of dollars in the construction of a large openai supercomputer to assist in powering OpenAI’s ChatGPT chatbot. Microsoft describes how it built Azure’s robust artificial intelligence infrastructure—which is utilized by OpenAI—in two blog articles that were released on Monday. It also discusses how its systems are becoming even more reliable.
Microsoft claims to have connected thousands of Nvidia graphics processing units (GPUs) on its Azure cloud computing platform to create the supercomputer that drives OpenAI’s initiatives. As a result, OpenAI was able to train ever-stronger models and “unlock the AI capabilities” of programs like Bing and ChatGPT.
According to a statement sent to Bloomberg, Scott Guthrie, vice president of AI and cloud at Microsoft, stated that the initiative cost the business several hundred million dollars. Even though Microsoft has extended its multiyear, multibillion-dollar investment in OpenAI, it may not seem like much. Still, it does show that the company is eager to pour even more money into the AI industry.
trip-wizards | runwayflight | travelsswiftly | globe-wander | tourepicness
With the introduction of its new virtual machines that utilize Nvidia’s H100 and A100 Tensor Core GPUs along with Quantum-2 InfiniBand networking, a project both companies hinted at last year, Microsoft chatgpt is already striving to make Azure’s AI capabilities even more potent. Microsoft claims that this should make it possible for OpenAI and other Azure-dependent businesses to train more sophisticated and sizable AI models.
Eric Boyd, corporate vice president of Azure AI at Microsoft, said in a statement, “We saw that we would need to build special purpose clusters focusing on enabling large training workloads and OpenAI was one of the early proof points for that.” “We closely collaborated with them to identify the essential elements they needed and the things they were looking for as they constructed their training environments.”
