Advanced Computing in the Age of AI | Tuesday, March 19, 2024

Google Nation: $10B for Office, Data Center Expansion in 11 States 

Google disclosed in a blog it will invest $10 billion in office and data center expansion around the U.S. this year, generating thousands of new jobs in the process. This follows an investment of $13 billion in its national footprint last year.

According to Sundar Pichai, Alphabet and Google CEO, Google will expand in 26 states, with new facilities in Texas, Washington, Pennsylvania, Ohio, Oklahoma, New York, Nebraska, Massachusetts, Georgia, California and Colorado.

“These investments will create thousands of jobs—including roles within Google, construction jobs in data centers and renewable energy facilities, and opportunities in local businesses in surrounding towns and communities,” Pichai said.

“This effort builds on the momentum of the $13 billion investment in communities from South Carolina to Nevada we made in 2019. Combined with other R&D investments, Google’s parent company Alphabet was the largest investor in the U.S. last year,” he said, citing confirmation in a report from the Progressive Policy institute.

In the south, Google plans to grow its engineering teams in its Atlanta offices while expanding offices and data centers in Texas, Alabama, South Carolina, Virginia and Tennessee. In Mississippi, the company plans to open a customer service operations center.

On the west coast, the company will expand its Google Cloud campus in Seattle and undertake a development in Kirkland, WA, that will open later this year. New offices and data centers also will be built in Oregon, in Los Angeles and in the San Francisco Bay area, along with a $1 billion distribution to fund housing there.

In the midwest, the company recently opened a Google Cloud office in Chicago and expanded in Madison, WI; it plans to open offices in Detroit, a data center in Ohio and finish expansion of its Iowa data center. In Colorado, Pichai said Google has the capacity to double its workforce, and it plans to expand its data centers in Nebraska and Oklahoma.

The company also touted programs that have trained more than 4 million Americans to “access new digital opportunities.”

The infrastructure in Google’s expanding data center footprint will include custom servers built to its specs run in part by custom processors of its own making. This is a growing trend among hyperscalers and other major tech companies whose AI workloads demand for compute power exceeds a Moore’s Law-class increase in throughput. In Google’s case, the company built the TPU (Tensor Processing Units) accelerator for neural network machine learning; Microsoft collaborates with UK chip developer Graphcore to use, and to offer, that company’s intelligent processing units (IPUs) for ML workloads on Azure; and public cloud leader AWS launched its Graviton2 chips at Amazon’s most recent re:Invent conference, stating that the Arm processor delivers up to 40 percent better performance for 20 lower cost than comparable x86-based cloud instances.

EnterpriseAI