Intel and Alphabet’s Google Cloud said on Wednesday that they have collaborated to create a new type of chip that Intel hopes will become a major seller in the booming cloud computing market.

This new chip, called Evans, will be sold to companies other than Google, reflecting the way cloud computing providers operate. They build huge data centers filled with powerful physical computers, and sell virtual slices of these machines to other companies, which in turn are more cost-effective than building their own machines.

For cloud providers, tasks such as setting up virtual machines and delivering customer data to the correct location are essentially indirect costs. The Evans chip, which Google and Intel call the “Infrastructure Processing Unit” (IPU), separates these tasks from the main computing tasks and accelerates them. Doing so also helps to ensure that these functions are from hacker attacks and increase the flexibility of the data center.

“We think this is strategically important. This is an extremely important for us and the data center,” Nick McKeown, senior vice president of Intel Networks and Edge Group, told Reuters.

Intel is not the only company manufacturing infrastructure chips. Nvidia and Marvell Technology offer similar but slightly different products.

But Intel and Google are working together to develop a set of freely released software tools, hoping to make Intel’s version of the chip a more widely used industry standard outside of Google’s data center.

Google researcher and vice president of Amin Vahdat said that Google hopes to inspire a technology trend that will make it easier for all data center operators to flexibly divide physical computer servers into virtual servers to adapt to any computing tasks.

See also  Cloud Advisor

“The basic question of what a server is will go the content inside the metal plate. IPU will play a central role there,” Vahdat told Reuters.

© Thomson Reuters 2021