(Reuters) - Hewlett Packard Enterprise ( HPE ) CEO Antonio Neri pitched a new line of hardware, software and services to businesses interested in deploying various forms of artificial intelligence at its Las Vegas conference on Tuesday.

Since the launch of OpenAI's ChatGPT in 2022, businesses of all sizes have been scrambling to develop and incorporate AI, to do things such as make further use of the vast amounts of data they store.

Yet the chips, hardware and software needed to build and run the necessary AI applications can often be complex and risky, according to Neri.

"These customers want simplicity," he said.

In HPE's view there are "tens" of companies building advanced AI models and over a hundred governments potentially interested in building supercomputers and cloud computing infrastructure. But there is a significant opportunity to sell advanced AI hardware to the large group of businesses that are not focused on pushing the frontiers of tech.

"Enterprises are in the hundreds of thousands, if not millions," Neri said in an interview with Reuters ahead of the conference.

At the moment, HPE's AI business for that segment is roughly 15% of its AI revenue, with the vast majority of the remaining amount going to large cloud computing companies such as Amazon.com ( AMZN ) and Microsoft ( MSFT ).

HPE said it will offer servers powered by Nvidia's ( NVDA ) advanced graphics processing units used for AI computations and a host of other hardware and software the company has built. The new family of server and services products will be available in the fall, the company said.

On top of the hardware and software, HPE said it planned to offer consulting services that included help to manage and maintain the systems.

"This announcement is about the customer segment, to speed up the deployment of AI to accelerate this industrial revolution," Neri said.