Google‘s problem clearly isn’t name recognition. But its previous attempts at making the jump from consumer to enterprise service have been spotty, to be generous.

Last year at its inaugural NEXT conference, held at an abandoned San Francisco pier, Google Cloud SVP Diane Greene promised that Google Cloud would double down on the enterprise. It was a statement of faith. A year later at Moscone Convention Center, Google Cloud is now the hot spot for recruiting at Alphabet; it is on a multi-year commitment to triple the technical resources they put in the field from solution architects and deployment specialists to a specialized group that they term reliability engineering.

Of course a decade ago, Amazon was known for retailing, not computing. There’s a natural comparison between the two — Google and Amazon could be considered twins separated at birth. Both carved leadership in different spheres of online commerce; Google’s AdWords was a money machine that could direct you to Amazon’s loss leader online store.

Beneath the covers, both relied on technology to dominate their respective parts of the internet. But that’s where the similarity ends. Amazon’s core competence, taking cost out of operation and the back office, eventually proved a natural draw for enterprises trying to take cost and inertia out of their computing operations. Google cultivated a reputation for cutting edge technology, from its vision of the No-ops data center to its liberal use of artificial intelligence in its core business. Google’s research provided the direction for Yahoo, Facebook, Twitter, and others to invent the core building blocks of Hadoop.

And it’s still making itself known for “cool” things, like announcement of a private beta for Cloud Video Intelligence API that will make entities inside videos searchable. But this year, there was also attention to the practical: a beta of a new Cloud Dataprep service that embeds Trifacta’s data wrangling into Google big data portfolio, and making commercial data sets accessible to Google; BigQuery.

A few years back, much of the innovation that Google researchers wrote about was considered blue sky. Today, even Amazon wants in on deep learning. Naturally, Google is seeking to one up it, with the acquisition of Kaggle being a stake toward developing an AI community tied more closely to the Google Cloud.

Google want to convince you that the advanced features of its cloud make it the more reliable one. It claims that the longest outage they had lasted about 18 minutes, which occurred several years ago; by contrast, Amazon’s most recent outage on February 28, which took S3 down in several regions, lasted roughly four hours.

Of course, reality isn’t black and white; while traffic on Google’s network is hardly trivial, as a public cloud, it handles only a fraction of the traffic. Furthermore, Google Cloud’s record hasn’t been that spotless — a significant Gmail outage hit the North America and the UK last fall.

But as to the enterprise message, the biggest difference this year is that Google Cloud has started filling in the white space. Last year at a customer panel for analysts at NEXT, the head of technology for a large soft drink company that used Google Cloud for an expansive 2014 World Cup promotion admitted under his breath that he wouldn’t have necessarily thought of Google for running their SAP applications. A year later, Google Cloud announced its first major enterprise software partner: SAP.

Admittedly, the SAP partnership for now is off to a modest start, with the agreement to offer the HANA in-memory database on the Google Cloud, just like it already does with AWS and Azure. But there is significant potential upside here. Metaphorically, SAP looks like one of a cast of thousands with other cloud players. But as one of the first enterprise software providers to publicly align with Google Cloud, there is excellent potential for joint branded offerings that could meld Google’s deep learning, globally scaled Spanner database, and NoOps to provide smarter, simpler, next-generation SAP applications powered by HANA — and the Google Cloud. It’s making one modest step in that direction with a unique offering of the freemium SAP HANA Express for developers on the Google Cloud.

The question is whether the Google Cloud will meet customers where they are, or require then to change to take advantage of Google’s unique cloud infrastructure.

On one hand, Google’s database stack for the most part aligns with what Amazon and Microsoft offer. On the other, under the hood, the Google Cloud operates uniquely. Inside the cloud, data is encrypted continuously, both in motion and at rest. All compute is “serverless” (meaning you don’t have to worry about where or how code will execute).

And if your organization wants to take advantage of Cloud Spanner, you will have to re-model your database as it is optimized for horizontal scale and therefore lacks relational database features such as foreign keys. Likewise, BigQuery is not your garden variety data warehouse. While it exposes to the work as SQL, it also supports nested fields and tables. You could implement your star, snowflake, or OLAP cubes as is in BigQuery, but to get the most mileage, you’ll want to look at how to use nesting as introducing new flexibility to the way you model data.

Having pioneered modern cloud computing a decade ago, Amazon transformed the notion of selling surplus compute cycles in its online retail back office to become the company’s not so hidden profit machine. With such a head start, it’s almost a miracle that Microsoft Azure has managed to claw its way to becoming the viable second source that mature markets demand.

But Google is betting that, despite the more than 90 percent penetration of AWS and Azure, that there’s still room for a third major player. A pretty audacious stretch goal, and we take it seriously because Google is so hugely capitalized that its cumulative $32 billion investment in cloud infrastructure has barely caused a dent in its share price.

With its eye on enterprise, Google revamps Hangouts:



Source link