Microsoft and Amazon are the only two cloud suppliers be designated "leaders" in the Gartner 2017 Magic Quadrant for Infrastructure as a Service. It remains to be seen whether Google or anyone else will eventually join them.
Google is the only cloud supplier with the prospect of doing so. IBM and Oracle remain still distant candidates for the foreseeable future.
Google is situated just below AWS and Microsoft in the visionaries' quadrant but almost touching the line separating the two. Is it about to break through or will it perpetually find the leader's wall too high to breach? Of all the contenders, Google's Cloud Platform, with its improvements to App Engine and Compute Engine, is trying hardest to get there.
Diane Greene, hired by Google in November 2015 as senior VP of Cloud Platform, brought the experience of leading VMware as CEO from start-up to the point where it became an enterprise fixture. Google is hoping some of that experience will rub off on the platform. In March at Google's Next show in San Francisco, Greene positioned Google Compute Cloud not simply as plain vanilla infrastructure as a service but as a platform from which crucial AI services would flow in the future.
What's hard to remember that Google has been a cloud vendor for a long time. Since 2008, it has had App Engine, but it was a platform as a service meant to give developers access to a Google-like operational environment. Google has been a general purpose, infrastructure as a service provider for a much shorter time with Compute Engine having become generally available in December 2013. So Google is a 4.5-year-old supplier of what the enterprise wants.
Want to learn more about Google cloud? See Google Cloud Polishes Its Appeal to the Enterprise.
This year's Gartner assessment provides some clues on how close Google is coming to the leaders' quadrant, though it's still hard to tell whether it's about to cross over or is just treading water. The analysts said Google is trying to outgrow it's early tendency to appeal to startups and those other companies who most want to run functions like Google Search and cloud operations.
Google is clearly a Web company that knows how to run at scale. It now offers a limited number of customers a Customer Reliability Engineering program to teach customers to run cloud operations the way Google site reliability engineers do.
Google has a second avenue of appeal in that it originated the Kubernetes container cluster orchestration project and is viewed as an expert in running containers at scale, an efficient form of operations that appeals to many enterprise IT staffs.
"Google's ability to see to a broad range of customers has improved significantly over the past year," a result of its deeper investment in the platform, the analysts wrote. It is increasingly competing for digital projects in the enterprise.
Google also has a more flexible approach to licensing instances than AWS and more precise billing. It charges by the minute as opposed to by the hour, and consequently doesn't round up to the next hour after 15 minutes of entering a new hour's use.
The analysts said Gartner clients typically choose Google "as a strategic alternative to AWS." Customers who both use AWS and compete with AWS havw business reasons for wanting an alternative, and many of them are more open-source oriented or DevOps-centric, which makes them a poorer fit for Microsoft Azure.
Google's core services "are solid and well implemented," the analysts said, but "its feature set and scope of services are not as broad as that of the market leaders." Google has focused on providing differentiated features based on BigQuery analytics and Cloud Spanner distributed database service.
"Google is introducing more capabilities and partnerships that are important to enterprise customers" but it only recently started to focus on IT processes and enterprise applications that reflect more than just the Web-company and Web-scale style of operations. Cloud-native approaches it's got; legacy enterprise application operations it's not.
"It still needs to invest deeply in global expansion; Google Cloud Platform has data centers in just five countries, though it intends to enter an additional eight countries during 2017," they said.
Until now, Google has lacked much of an ecosystem around its IaaS capabilities, and what ecosystem there is tends to be focused on Google's platform as a service with its Java and Python orientation. "It has few managed service provider and infrastructure-centric professional services partners," although its own technical support is highly competent, they said.
Google acquired Orbitera last August for $100 million to establish a multi-cloud marketplace, which should help bring more independent software vendors to its platform. It's also trying to build out a set of platform management tools and recruit partners. Even if it succeeds, it will take time for the partners to build out their capabilities for the platform, they wrote.
The analysts left the issue hanging. While Google appears to be making the right moves, it's also appeared that way in the past and Google still doesn't have a lot to show for its high profile hiring of a new cloud platform senior vice president. Its pace of innovation and feature addition is increasing. Will it come up with some that make Google Cloud Platform a compelling choice in enough enterprise customers' decision-making?
IBM and its startup/gamer base
In a somewhat similar manner, the Gartner analysts offered a mixed assessment of IBM IaaS. IBM offered its Bluemix cloud as a platform as a service or developers before 2013, then acquired the multiple, infrastructure centers of SoftLayer in that year. SoftLayers centers, in addition to the U.S., are located in 15 countries. By the end of 2016, IBM was able to offer SoftLayer IaaS services through its Bluemix centers as well. SoftLayer is different from AWS in offering bare metal servers for use by single tenants as well as multi-tenant, virtual servers.
Despite IBM's mainframe history and number of legacy customers, its cloud operations have appealed more to start-ups and gaming companies, given SoftLayer's user interface's emphasis on small and medium business during it's own development. Gaming hosts also like its offering of bare metal servers.
Like AWS Lambda embedded functions and Microsoft Azure Functions, IBM's Bluxmix offers its own vehicle for serverless computing in OpenWhisk.
In general, the IBM SoftLayer cloud is good for general business applications, where API control over scalable infrastructure is sought and bare metal servers for both performance and regulatory compliance. IBM is currently engaged in redesigning the way its SoflLayer operation performs with customers, the analysts said. It is most likely seeking to give it a greater economies of scale and a better enterprise fit and finish, but IBM has made no announcement of when it will be ready. "The eventual rollout of its next-generation infrastructure architecture is likely to help IBM evolve beyond its current status as a hosting-scale provider." As such, it can't quite "match the cost economics of the market leaders," the report said.
In the meantime, SoftLayer's feature set "has not improved significantly since the IBM acquisition in mid-2013," meaning it is small and medium business oriented and missing features required by enterprise cloud customers, according to Gartner. It introduced an OpenStack-based infrastructure design via the Bluemix portal in 2016, then discontinued the attempt. "Customers must thus absorb the risk of an uncertain roadmap," and continued uncertainty over where IBM is placing its bets means a partner ecosystem that has stalled, the analysts wrote.
Bluemix is hosted in only three SoftLayer data centers and is not available in a nearby geography to many SoftLayer customers. Bluemix and SoftLayer do not share a single self-service portal and catalog with a consistent command line interface and API. Customers can't achieve a single view across the two environments or get a unified security context across the two.
The next generation infrastructure offering may change this but until IBM gives it a release date and makes an outline of its features known, it's likely to remain distant from the "leaders" in the Gartner Infrastructure as a Service quadrant.
Dell integrates EMC units
Dell Technologies is now a cloud supplier, after it completed its merger with EMC, VMware and Pivotal Software last September. EMC had acquired Virtustream, an infrastructure as a service supplier, and appeared at the top of the heap among the eight players occupying the Gartner's "niche players" quadrant.
Virtustream was founded in 2008 and acquired by EMC in 2015. It has cloud data centers in the Eastern and Western U.S., the U.K., France, Germany, the Netherlands, Australia and Japan. It's focus has been to provide infrastructure for traditional business applications and be usable by traditional IT staffs, "especially those seeking to improve their agility," the analysts wrote.
In fact, Virtustream was unique in seeking to provide cloud services for running complex enterprise application suites, such as ERP from SAP, Oracle and Blackbaud. It supplies an application platform designed for high availability, security, governance and performance with an ability to meet specific SLA requirements.
Despite the turmoil of being acquired twice within a two-year period, Virtustream is "thriving under Dell… It has continued to win large-scale enterprise deals…" said the Gartner analysts. It can also provide managed application services for an SAP suite, the authors pointed out. It continues to expand the locations in which it is available and the number of complex workloads it can host for customers.
The writers cautioned, however, that Virtustream is not trying to provide broad, general purpose IaaS, like Microsoft and AWS are. Part of its success is in targeting complex application implementations that prompt a customer to purchase its professional services and use managed services on an ongoing basis.
"Virtustream's roadmap is inextricably tied to into other Dell entities," such as VMware, EMC and Pivotal, the report said. It is better suited for a complex application implementation "where an environment will be carefully and consultatively tuned" rather than a general purpose IaaS environment. Customers also need to define their role in operations versus those of the provider's staff, they said.
Oracle arrives in the quadrant
Oracle was added to the Magic Quadrant this year with its second generation IaaS offering, launched last November. It offers both KVM virtual machines and bare metal servers, plus ease of activation of the Oracle database and Exadata appliances.
Oracle has second generation cloud centers in the U.S. East and West. It operates earlier generation cloud data centers in the central and Eastern U.S., and in the U.K. and the Netherlands. There's the usual discussion of branding. What Oracle calls its Bare Metal Cloud Services "is really both a bare-metal and virtual cloud IaaS platform, the authors said.
Oracle has recruited "a highly experienced engineering team from hyperscale cloud providers" to build out its second generation offering. Oracle has "a realistic perspective on its late entry into the market and has a sensible engineering roadmap focused on building a set of core capabilities that will eventually make it attractive for targeted use cases," the writers said.
Regardless of whether Oracle will be able to directly compete for IaaS business, its Gen 2 offering "will be important for Oracle's PaaS and SaaS businesses." Once low-latency connections to other clouds become available, customers may want to run their Oracle databases on bare metal servers in the Oracle cloud, even when much of their infrastructure resides on an Azure or AWS, they said.
The authors warned that the Oracle Gen 2 cloud still has a limited operational track record and customers are dependent on Oracle support. There are few third parties available for this offering. "Customers need to have a very high tolerance for risk, along with strong technical acumen," the analysts advised.