Google on Tuesday upped BigQuery volumes and cut costs. It’s the latest in a series of moves designed to bolster its Cloud Platform and establish Google as the go-to tool for storing, computing and analyzing large data sets.
One big change is that BigQuery can now run more statistical calculations on data, such as percentiles and ranking of query results. What’s more, BigQuery will return larger data sets as a result of queries. Previously, there was a limit of 128MB of compressed data; now that is no more, in response to customer requests, according to a Google blog post from BigQuery Product Manager Ju-Kay Kwek and another one from Cloud Platform team member Felipe Hoffa.
Also, query results can now stay cached for a day, so analysts looking to just read results that stay the same can save some money. An updated interface lets users know right away if the syntax of their queries contains errors, and if everything is kosher, it will provide an estimate for the cost of the query.
Additionally, Google is doubling the size limits on queries. More detail is available from Hoffa’s blog post.
And on July 1, prices will drop from 12 cents per gigabyte per month down to 8 cents per gigabyte per month. The biggest users will apparently be eligible in the future to sign up for tiered query pricing, too.
Infrastructure-as-a-Service (IaaS) providers from Microsoft to Rackspace have become quiet adept at cutting prices and adding features, as competition keeps heating up. Market leader Amazon Web Services is arguably the king of price cuts, and in fact on Monday it struck again by slashing the prices of its Relational Database Service (RDS) instances.
Google has lowered the prices of its own cloud services before, too, on Cloud Datastore, Google Cloud Storage and Google Compute Engine instances. BigQuery can be considered more of a Software as a Service play, although like GCE and others, it’s part of the Google’s broader IaaS strategy.
BigQuery is already a fast solution for running SQL-style interactive queries on large data sets. Speaking at GigaOM’s Structure: Data conference last year, Kwek (pictured) played up BigQuery as an analytics tool offering much faster speeds than what’s possible on premise, without having to store that data on site. (Don’t miss GigaOM’s Structure conference next week, at which Google Fellow and MapReduce paper co-author Jeff Dean will make an appearance.) With the price cuts and new features, BigQuery looks more palatable, and that boosts the prospects of the larger Cloud Platform.