Oracle Cloud Puts Data Center Expansion Pedal to the Metal

After years of watching its largest rivals in cloud services spend billions of dollars on infrastructure quarterly – while keeping its own data center investment relatively modest – Oracle’s cloud business is accelerating expansion.

The company announced on Monday a plan to add 20 new cloud availability regions to the existing 16 by the end of next year. The plan includes new Oracle cloud data centers in countries across the Americas, Europe, Middle East, and Asia-Pacific.

Unlike its largest rivals Amazon, Microsoft, and Google, who both design and build their own data centers in addition to leasing facilities from specialist developers, Oracle is planning to continue relying on the latter exclusively, Leo Leung, senior director of product and strategy for Oracle Cloud Infrastructure, told Data Center Knowledge in an interview.

If all goes as planned, there will be 26 Oracle Cloud availability regions by the end of 2020. There are currently 22 Amazon Web Services availability regions, 54 Microsoft Azure regions, and 20 Google Cloud Platform regions.

Oracle’s accelerated expansion will provide a substantial business boost to what will likely be several specialist data center developers, the likes of Digital Realty Trust, Equinix, or Global Switch. Depending on the market, rental rates for data center leases over 250kW this year have been anywhere between $80 and $260 per kilowatt per month, according to a recent report by the commercial real estate services company JLL.

Oracle didn’t say how much capacity it would deploy in each of the upcoming regions, but each availability region consists of multiple data centers, totaling several megawatts of power capacity. Using JLL’s range, a 1MW lease generates $960,000 to $3 million in annual revenue for a data center provider. That doesn’t include various ancillary services the provider may offer its clients.

While Oracle’s cloud business has been growing, it’s been doing so slowly compared to the double-digit leaps in revenue growth AWS and Azure, the two cloud market leaders, have been reporting quarter after quarter. At the same time, revenue from Oracle’s bread-and-butter enterprise software sales has been shrinking, raising the question of whether the company will ultimately be able to accelerate cloud growth quickly enough to outpace the rate of decline in its other businesses.

Oracle announced the acceleration of its cloud data center footprint expansion at OpenWorld, its big annual conference in San Francisco. The event, taking place this week, will be the first in years without an appearance by Mark Hurd, Oracle’s co-CEO, salesman-in-chief, and major force behind its cloud business growth. The company announced last week that Hurd has taken a leave of absence for an undetermined length of time due to a health issue but did not share any more details. Safra Catz, the second co-CEO, and Larry Ellison, Oracle’s co-founder, CTO, and executive chairman, will share Hurd’s responsibilities, the company said.

Also on Monday, Oracle Cloud announced an expansion of its recent partnership with Microsoft Azure and touted big new customers, including Cisco and McAfee, both of whom, according to Leung, have new software-as-a-service security products running on Oracle’s cloud infrastructure.

Oracle announced its own new cloud security services at the show, rolled out a new free tier that gives developers access to a limited amount of Oracle Cloud resources for an unlimited amount of time, and launched Autonomous Linux, which automatically keeps clients’ Linux instances running in Oracle Cloud patched and up to date. It also launched a management service for Autonomous Linux, non-autonomous Linux, and Windows. Leung explained that the Autonomous Linux feature is meant to be used primarily together with the OS management service.

Additionally, Oracle announced that the latest, second-generation Exadata, the hardware system designed for maximum Oracle database performance, is now available as part of Oracle Cloud at Customer, the portfolio of cloud services the company provides from especially security- and compliance-sensitive customers’ own data centers.

One New Region Every 22 Days

Today, Oracle Cloud offers 11 availability regions for civilians and five for government customers, including locations in the US, Canada, Brazil, Germany, the UK, Japan, South Korea, India, and Australia. Between now and the end of next year, the company expects to launch one new region every 22 days on average. That’s faster than Oracle, or any other cloud provider, has ever added new availability regions, Leung claimed.

New Oracle Cloud availability regions due to come online by the end of 2020:

Besides services for its cloud customers, the Oracle Cloud Infrastructure platform also supports Oracle’s own Software-as-a-Service products, a portfolio collectively called Oracle Fusion. These include things like enterprise resource planning (ERP), customer relations management (CRM), and human capital management (CHM) applications sold as subscriptions rather than traditional software licenses. The Fusion services will be available from all the new regions, Leung said.

Expanding Azure Multi-Cloud Partnership

Hoping to accelerate the growth of its cloud business, Oracle earlier this year struck a partnership with Microsoft Azure, one of its two biggest rivals in the market, to make it easier for customers to run applications on Azure using their Oracle databases running on Oracle Cloud. The companies said they would connect the two cloud platforms by making their networks and identity management interoperable.

At the time of the announcement (in June), this multi-cloud capability was available in Ashburn, Virginia; and London. On Monday, Oracle said it would expand it to availability regions in the US West, Asia, and Europe.