Fresh off its gadget-packed press conference on Wednesday, Google LLC is shifting focus to the enterprise with a major cloud update.

In a blog post published today, the company introduced a new class of virtual machines for the Google Cloud Platform. The series is based on Intel’s Skylake chip architecture and offers 96 cores with every instance. That’s more than any other competing offering that runs on the processor family, according to Google.

The company is offering three instance configurations that come with 60 to 624 gigabytes of memory. Not too surprisingly, Google’s main target market appears to be analytics workloads.

The company has certified the 624-gigabyte edition for SAP SE’s HANA in-memory store and BW/4HANA data warehouse. The two products are used extensively among large enterprises to process relational records such as those generated by back-office applications. With that said, there is no shortage of competing software, which means that Google may very well add support for other products in the future.

Companies looking to process particularly large data troves can combine up to 16 of the 624-gigabyte instances into a cluster with 9.75 terabytes of memory. And as if that weren’t enough, Google is working on even bigger virtual machines that will provide up to 4 terabytes of memory each.

It’s not difficult to see why the technology giant is so keen on accommodating data-intensive applications. Enterprise analytics projects tend to produce large volumes of information that need to be stored, managed and processed, which can add up to a lot of revenue for a cloud provider. That’s why rivals such as Amazon Web Services Inc. have their own selection of analytics-optimized instances.

Google and its competitors also offer managed services designed to ease other aspects of data processing. The search giant’s efforts on this front are bearing fruit: It was recently named by Forrester Research Inc. as the leader of the so-called “Insight Platform as a Service” segment.

Image: Theerawat Sangprakarn