Google this week shored up its case for being viewed as a major public cloud service provider (CSP) via a bevy of updates designed to make Google Cloud Platform (GCP) more broadly applicable to enterprise IT organizations.
New offerings launched at the Google Cloud Next 2017 conference included a beta release of an implementation of the open source Postgres database that is compatible with Oracle applications alongside the general availability of an instance of Microsoft SQL Server that also supports Microsoft.Net applications.
Those offerings, however, will vie for attention among enterprise IT organizations against a recently launched Google Spanner database service based on a time-stamped architecture tied to atomic clocks, which Google touts as providing all the benefits of a SQL and ACID compliant transaction processing environment that scales out as simply as a NoSQL database.
Google also revealed it is making available 64-core virtual machines with up to 416 GB of memory and that it will be the first CSP to make use of Intel Xeon Skylake processors. Google also pledged to make as much as 1TB or memory and even higher core counts available later this year.
Speaking at the conference, Urs Hölzle, the senior vice president of technical infrastructure at Google, says one of the things that distinguishes GCP most is the level of flexibility it provides in terms of how virtual machines are consumed. Rather than requiring customers to commit to a specific size of virtual machines or a length of time to employ them, which ends in organizations wasting nearly half the compute resources they have paid for in the cloud, Hölzle says GCP allows public cloud services to be consumed in a more granular fashion using sustained and committed use discounts that make realizing the true promise of cloud computing simpler, at a total cost that is 60 percent less than rival platforms.
“Only GCP is a truly elastic cloud,” says Hölzle. “Cost and capacity planning should be a distant memory.”
As part of the effort to make public clouds more elastic, Google this week also extended it serverless computing service to include a beta release of Cloud Functions, which is integrated with the Firebase platform for building mobile applications.
Also on tap this week are new pipeline tools for use in the Google BigQuery data warehouse as well as data preparation and application programming interfaces (APIs) for data loss prevention to make it simpler to manage data as well as redact personally identifiable information (PII).
Finally, Google announced it will make available a chip to better secure access to its cloud services alongside a beta release of an Identity Aware Proxy service, and is making a Google Cloud Key Management Service generally available.
GCP right now, in terms of being a public CSP, still runs a distant third to AWS and Microsoft Azure in adoption. But as Google figures out how to better address enterprise application requirements with help from partners such as SAP, the gap between GCP and the two leading CSPs should begin to narrow.
