Intel, Google Co-Designed Data Centre E2000 Chip

data centre google

Google Cloud deploys its new C3 machine series, featuring the Intel infrastructure processing unit E2000, co-designed by Google and Intel

Google Cloud and Intel have revealed their data centre chip called the E2000, co-designed by both firms.

Intel announced the development on Tuesday calling it “significant milestone in delivering an end-to-end programmable platform.”

Google Cloud has apparently deployed its new C3 machine series, which features Intel Xeon processors in private preview and debuts the Intel infrastructure processing unit E2000 (codenamed Mount Evans), which was co-designed by Google and Intel.

PICTURE: Programmable infrastructure processing units offer Intel partners and customers improved security, reduced overhead and opportunities to free up CPU performance. (Credit: Intel Corporation)

E2000 chip

Intel claims that thanks to its unique architectural approach, “C3 machine instances deliver strong performance gains of up to 20 percent over previous generation C2 instances, enabling high performance computing and data-intensive workloads.”

It said that Google Cloud C3 machines also pave the way for a future where infrastructure processing units (IPUs) are integrated into data centres, accelerating cloud infrastructures and maximizing performance.

“We are pleased to have codesigned the first ASIC infrastructure processing unit with Google Cloud, which has now launched in the new C3 machine series,” said Nick McKeown, Intel senior VP, Intel Fellow and general manager of Network and Edge Group.

“A first of its kind in any public cloud, C3 VMs will run workloads on 4th Gen Intel Xeon Scalable processors while they free up programmable packet processing to the IPUs securely at line rates of 200 gigabits per second,” McKeown added.

“This Intel and Google collaboration enables customers through infrastructure that is more secure, flexible and performant,” said McKeown.

Google Cloud has detailed its C3 machine series here.

Chip developments

This is not the first time that Google has lent its expertise to chip designs.

Google historically has been notoriously secretive when it came to discussing the technology it uses and indeed even the design of its data centres.

But the firm has long experimented with its own chips.

In 2013 for example it was reported it was considering developing its own range of server microprocessors, and even mooted the possible development of smartphone chips.

And in 2016 Google revealed that it was making its very own chips, designed specifically for machine learning or artificial intelligence (AI).

Then in August 2017 details began to emerge of Google’s security chip (Titan) for its Google Cloud Platform (GCP), which was predominantly a hardware protection device.

In July 2022 Google Cloud announced its first ARM-based VMs (virtual machine) utilising chips from US server chip firm Ampere Computing.

Google at the time said some customers are looking for ARM architectured options to deliver excellent performance per watt efficiency.

Both AWS and Azure already offer ARM-based VMs.

But more recently the US Commerce Department and Google in September this year reached a co-operative research and development agreement, to produce chips that researchers and startups can use to develop new nanotechnology and semiconductor devices.

Google will essentially finance the initial cost of setting up production and will subsidise the first production run of the chips, which will be manufactured by SkyWater Technology at its Bloomington, Minnesota, semiconductor foundry.