Will Google Design its Own Processors?

By

Though some think the idea is a leaked negotiating ploy, there’s a rumor Google might design its own server processors. The idea might seem farfetched, but Google now builds many related products it might otherwise buy.

Operating systems, browsers, PCs, tablets, smart phones, global backbone networks, access networks and data centers are among the inputs and outputs Google now directly supplies, when it might simply source them externally from third parties. That is one reason any rumor about moving to customer processors does not seem wildly out of place.

According to the rumor, Google is considering designing its own server processors using technology from ARM Holdings. In principle, by using its own designs, Google could better manage the interactions between hardware and software.

Any such move could prompt other major application providers, who also rely on their own data centers to support cloud operations, to make their own moves, with obvious implications for Intel, which supplies most of the server processors used in such devices.

Google alone is Intel’s fifth largest customer, according to Bloomberg, buying at least $500 million worth of Intel server chips each year.

As much as Intel would have reason to fear such a move, rival ARM would obviously gain, particularly since, though ARM is a major supplier in mobile devices, ARM has small presence in the server end of the business.

Intel, conversely, has little share in the mobile devices market, where low power consumption is a big issue.

Google in August 2013 joined OpenPOWER, a new initiative under which IBM will license the core intellectual property for POWER technologies to other companies for use in designing servers employed in cloud data centers.

OpenPOWER makes it possible for cloud services and their technology providers to redesign processors and circuit boards to optimize the interactions of microprocessors, memory, networking, data storage and other components.

As a result, they can get servers that are custom-tuned for their applications. That would suggest there is something to the rumor, despite legitimate skepticism.




Edited by Cassandra Tucker
Get stories like this delivered straight to your inbox. [Free eNews Subscription]

Contributing Editor

SHARE THIS ARTICLE
Related Articles

How Is AI Bringing Intelligence to Traditionally Manual Inspection Workflows?

By: Contributing Writer    10/16/2026

For insurers, rental companies, fleet operators, and remarketing businesses, vehicle inspections are a part of everyday operations. Much of that proce…

Read More

3 Factors That Can Dramatically Impact Your Microsoft Intune Spend

By: Contributing Writer    10/13/2026

With remote and dispersed work here to stay for the long term, secure endpoint management has become a big headache for IT and cybersecurity teams. Th…

Read More

Why Threat Modeling Fails Without an Attacker's Perspective

By: Contributing Writer    10/13/2026

A security team might spend weeks mapping attack paths and planning defenses, only to be breached through a web application nobody remembered was stil…

Read More

From AI Experiment to Enterprise Strategy: Making Claude Work at Scale

By: Contributing Writer    10/12/2026

Artificial intelligence often enters an organization through experimentation. One employee discovers a useful prompt, another tests document analysis,…

Read More

SIEM needs exposure context to keep pace with AI-era threats

By: Contributing Writer    10/6/2026

Security information and event management (SIEM) platforms give security operations center (SOC) teams a central place to collect and analyze security…

Read More