In the 2017 Magic Quadrant for Cloud Infrastructure as a Service, Worldwide, for the 7th straight year, Gartner placed Amazon Web Services in the “Leaders” quadrant and positioned the furthest on completeness of vision and the highest ability to execute axes.
The market for cloud IaaS is dominated by two leading service providers. Other service providers have responded by launching new offerings, but customers must carefully manage the risks of adopting less-mature offerings.
Cloud computing is a style of computing in which scalable and elastic IT-enabled capabilities are delivered as a service using internet technologies. Cloud infrastructure as a service (IaaS) is a type of cloud computing service; it parallels the infrastructure and data center initiatives of IT. Cloud compute IaaS constitutes the largest segment of this market (the broader IaaS market also includes cloud storage and cloud printing). Only cloud compute IaaS is evaluated in this Magic Quadrant; it does not cover service providers that exclusively offer cloud storage, platform as a service (PaaS), SaaS, cloud service brokerage (CSB) or any other type of cloud service, nor does it cover the hardware and software vendors that may be used to build cloud infrastructure. Furthermore, this Magic Quadrant is not an evaluation of the broad, generalized cloud computing strategies of the companies profiled.
Amazon Web Services
Amazon Web Services (AWS), a subsidiary of Amazon, is a cloud-focused service provider. It pioneered the cloud IaaS market in 2006.
Offerings: AWS offers Xen-virtualized multitenant and single-tenant compute (Elastic Compute Cloud [EC2]), with multitenant storage, along with extensive additional IaaS and PaaS capabilities, including object storage with an integrated CDN (Amazon Simple Storage Service [S3] and CloudFront), a Docker container service (EC2 Container Service [ECS]), a batch computing service (AWS Batch), event-driven “serverless computing” (Lambda), and an aPaaS-like developer experience (Elastic Beanstalk). It is willing to negotiate large-scale single-tenant and on-premises deals (such as the U.S. intelligence community cloud deal). The AWS Marketplace has an extensive selection of third-party software and services. Enterprise-grade support is extra. It has a multi-fault-domain SLA. Colocation needs are met via partner exchanges (AWS Direct Connect).
Locations: AWS groups its data centers into regions, each of which contains at least two availability zones (data centers). It has regions on the East and West Coasts of the U.S., and in Canada, Germany, Ireland, U.K., Australia, India, Japan, Singapore, South Korea and Brazil. It also has one region dedicated to the U.S. federal government. There is a China region operated by Sinnet, which requires a China-specific AWS account. It has a global sales presence. The portal and documentation are provided in English, Dutch, French, German, Italian, Japanese, Korean, Mandarin, Portuguese and Spanish. The primary languages for support are English, Japanese and Mandarin, but AWS will contractually commit to providing support in a large number of other languages.
Provider maturity: Tier 1. AWS has been the market share leader in cloud IaaS for over 10 years.
Recommended mode: AWS strongly appeals to Mode 2 buyers, but is also frequently chosen for Mode 1 needs. AWS is the provider most commonly chosen for strategic, organizationwide adoption. Transformation efforts are best undertaken in conjunction with an SI.
Recommended uses: All use cases that run well in a virtualized environment. Applications that are potentially challenging to virtualize or run in a multitenant environment — including highly secure applications, strictly compliant or complex enterprise applications (such as SAP business applications) — require special attention to architecture.
- AWS remains the dominant market leader, not only in IaaS, but also in integrated IaaS+PaaS, with an end-of-2016 revenue run rate of more than $14 billion. It continues to be the thought leader and the reference point for all competitors, with an accelerating pace of innovation on top of an already rich portfolio of services, and an expanding impact across a range of IT markets. It is the provider most commonly chosen for strategic adoption; many enterprise customers now spend over $5 million annually, and a few spend over $100 million. While not the ideal fit for every need, it has become the “safe choice” in this market, appealing to customers that desire the broadest range of capabilities and long-term market leadership.
- AWS is the most mature, enterprise-ready provider, with the deepest capabilities for governing a large number of users and resources. Thus, it is the provider not only chosen by customers that value innovation and are implementing digital business projects, but also preferred by customers that are migrating traditional data centers to cloud IaaS. It can readily support mission-critical production applications, as well as the implementation of highly secure and compliant solutions. Implementation, migration and management are significantly eased by AWS’s ecosystem of more than 2,000 consulting partners that offer managed and professional services. AWS has the broadest cloud IaaS provider ecosystem of ISVs, which ensures that customers are able to obtain support and licenses for most commercial software, as well as obtain software and SaaS solutions that are preintegrated with AWS.
- AWS’s extensive portfolio of services requires expertise to implement. This is somewhat mitigated by AWS’s excellent business-class technical support, accurate documentation, extensive training and certification, and a partner badging system that includes an Audited MSP Partner designation that helps customers choose experienced and capable MSPs. However, customers should be aware that while it’s easy to get started, optimal use — especially keeping up with new service innovations and best practices, and managing costs — may challenge even highly agile, expert IT organizations, including AWS partners.
- AWS has just begun to adapt to the emergence of meaningful competitors. AWS is perceived as a cost leader, and is the key reference point for pricing in this market, but it is not eager to be the lowest-cost bidder in a competitive situation. Its granular pricing structure is complex; use of third-party cost management tools is highly recommended. Its disciplined approach to contract negotiation and discounts is based almost solely on customer spending and near-term revenue opportunity. Although its baseline enterprise agreement (EA) has recently improved and offers very competitive T&Cs even prior to negotiation, only a few of its services are covered by an SLA.
Google is an internet-centric provider of technology and services. Google has had an aPaaS offering since 2008, but did not enter the cloud IaaS market until Google Compute Engine was launched in June 2012 (with general availability in December 2013).
Offerings: Google Cloud Platform (GCP) combines an IaaS offering (Compute Engine), an aPaaS offering (App Engine) and a range of complementary IaaS and PaaS capabilities, including object storage, a Docker container service (Container Engine) and event-driven “serverless computing” (Google Cloud Functions, in beta). Compute Engine VMs are KVM-virtualized and metered by the minute. Enterprise-grade support is extra. It has a multi-fault-domain SLA. Colocation needs are met via partner exchanges (Google Cloud Interconnect).
Locations: Google groups its IaaS data centers into regions, each of which contains at least two zones (data centers). There are East Coast, West Coast, and central U.S. regions, as well as regions in Belgium, Japan, Singapore and Taiwan. Google has a global sales presence. Support is available in English and Japanese. The portal is available in English, Dutch, French, German, Italian, Polish, Spanish, Turkish, Russian, Portuguese, Korean, Japanese, Mandarin and Thai. Documentation is available in English and Japanese.
Provider maturity: Tier 1. Gartner estimates that GCP is a distant third in cloud IaaS and integrated IaaS+PaaS market share. Google has massive investments in infrastructure for Google as a whole.
Recommended mode: GCP primarily appeals to Mode 2 buyers.
Recommended uses: Big data applications, batch computing and cloud-native applications.
- Google’s strategy for GCP centers on commercializing the internal innovative technology capabilities that Google has developed to run its consumer business at scale, and making them available as services that other companies can purchase. GCP is most attractive to cloud-native companies and those that want to “run like Google,” but it is now trying to win customers with traditional workloads and IT processes as well. Its Customer Reliability Engineering program (currently offered directly to a limited number of customers, as well as in conjunction with Pivotal) uses a shared-operations approach to teach customers to run operations the way that Google’s site reliability engineers do. Google is increasingly positioning itself as an “open” provider, and emphasizing portability; it is involved in many open-source ecosystems, including that of Kubernetes, its open-source container cluster management software.
- Google’s ability to sell to a broad range of customers has improved significantly over the past year — a visible impact from recent deeper investments in GCP. Google has become much more aggressive in its go-to-market strategy and is increasingly competing for digital business projects in enterprises. Google uses deep discounts and exceptionally flexible contracts to try to win projects from customers that are currently spending significant sums of money with cloud competitors. Gartner clients typically choose GCP as a secondary provider rather than a strategic provider, though GCP is increasingly chosen as a strategic alternative to AWS by customers whose businesses compete with Amazon, and that are more open-source-centric or DevOps-centric, and thus are less well-aligned to Microsoft Azure.
- GCP has a solid and well-implemented core of fundamental IaaS and PaaS capabilities, but its feature set and scope of services are not as broad as that of the market leaders. GCP is working on trying to achieve feature parity, but is also pursuing innovative and differentiated platform capabilities, such as its BigQuery analytics and Cloud Spanner distributed database, and its feature velocity continues to accelerate. Google is introducing more capabilities and partnerships that are important to enterprise customers, but until recently, it was highly focused on cloud-native applications and DevOps-style operations, not on applications and IT processes from the pre-IaaS era. It still needs to invest deeply in global expansion; GCP has data centers in just five countries, though it intends to enter an additional eight countries during 2017.
- Google has only recently begun to build an ecosystem around its IaaS capabilities, which significantly heightens the challenges of adopting GCP. The initial GCP ecosystem has been primarily focused on solution development and GCP’s PaaS capabilities. It has few MSP and infrastructure-centric professional services partners, although it has competent technical support staffed by software engineers. Google’s acquisition of Orbitera has given it a multicloud marketplace, and will help accelerate its efforts to bring ISVs to its platform. GCP has also been making aggressive efforts to build an ecosystem of management tools. It is actively recruiting partners, and prospective partner interest is high, but it will take time for such partners to build out capabilities.
IBM is a large, diversified technology company with a range of cloud-related products and services. In July 2013, it acquired SoftLayer, an independent web hoster with a focus on small or midsize businesses (SMBs), and in January 2014, it shut down its own SmartCloud Enterprise cloud IaaS offering, after migrating its existing customers to SoftLayer. IBM began to absorb the operations of SoftLayer, an IBM company, during 2016, and that process is ongoing.
IBM has two portals for cloud IaaS. The SoftLayer portal contains the full range of services that have previously been sold under the SoftLayer brand. However, in late 2016, IBM began to offer a subset of SoftLayer services through the service catalog in the Bluemix portal. Bluemix was originally IBM’s PaaS offering and has since expanded into a broader platform for IBM Cloud; in this context, the SoftLayer services are branded IBM Bluemix infrastructure. The Bluemix portal also contains some infrastructure services that are only available through the Bluemix service catalog. This Magic Quadrant evaluation considers the customer experience through both portals. We use “SoftLayer infrastructure” to refer to all SoftLayer services (whether cloud or noncloud), regardless of which portal is used to provision and manage them.
Offerings: IBM offers both multitenant and single-tenant Citrix-XenServer-virtualized compute (Virtual Servers), as well as paid-by-the-hour nonvirtualized dedicated servers (Bare Metal Servers). It has OpenStack-based object storage with an integrated CDN (via a partnership with Verizon Digital Media Services, formerly EdgeCast); the Bluemix portal additionally offers S3-compatible object storage based on Cleversafe technology. SoftLayer also has noncloud offerings, such as paid-by-the-month dedicated servers (a broader range of configurations than is available per hour) and hosted appliances, but IBM does not make a clear distinction between these offerings and its cloud IaaS capabilities. Bluemix has a Docker-based container service, event-driven “serverless computing” (OpenWhisk), a Cloud Foundry-based aPaaS, and other PaaS capabilities. Managed services are optional. Colocation needs are met via partner exchanges (IBM Direct Link).
Locations: SoftLayer infrastructure is located in multiple data centers in the U.S., along with data centers in Canada, Mexico, Brazil, France, Germany, Italy, the U.K., the Netherlands, Norway, Australia, Hong Kong, India, Japan, Korea and Singapore. IBM has a global sales presence. It offers support in the wide range of languages in which IBM does business. The portal and documentation are available in English, French, German, Italian, Portuguese, Spanish, Cantonese, Mandarin, Korean and Japanese.
Provider maturity: Tier 2A. IBM’s cloud infrastructure strategy has shifted over time. It has made multiple forays into the cloud IaaS market, and is currently building a new cloud IaaS offering.
Recommended mode: Before the IBM acquisition, SoftLayer typically sold to Mode 2 customers (specifically startups and gaming companies with a strong interest in bare-metal dedicated hosting). Since the acquisition, IBM has increasingly focused on acquiring Mode 1 customers, but SoftLayer infrastructure better meets the needs of Mode 2 customers (as long as they only require basic cloud IaaS and specifically want bare metal).
Recommended uses: E-business hosting, general business applications and batch computing, in circumstances that require both API control over scalable infrastructure and bare-metal servers in order to meet requirements for performance, regulatory compliance or software licensing. Alternatively, IBM outsourcing deals that use bare-metal servers as the hosting platform, where the customer has a need for supplemental basic cloud IaaS. SoftLayer infrastructure may also be used as a component of applications built using the Bluemix PaaS capabilities.
- IBM is in the midst of a “Next-Generation Infrastructure” (NGI) engineering project, but it has not announced a release date. IBM has set aggressive design goals for the performance and price point of this new cloud infrastructure, which adopts the principles of hyperscale infrastructure architecture. It uses a fully custom IBM integrated system to deliver software-defined infrastructure; it incorporates IBM hardware design innovations as well as custom software. It represents a significant step forward in IBM infrastructure capabilities as well as in IBM’s ability to serve the needs of future cloud-native applications, particularly in relation to IBM’s broader ambitions in cognitive computing. NGI will eventually be the basis of its future cloud IaaS offerings, as well as the platform for other IBM Cloud offerings, and will be presented as a Bluemix experience. IBM intends to gradually transition its existing infrastructure customers to NGI.
- IBM has a strong brand and existing customer relationships across the globe. IBM’s base of strategic outsourcing customers help drive cloud-enabled data center outsourcing business into SoftLayer data centers. Its developer ecosystem may help to drive adoption of Bluemix Infrastructure services. IBM intends to make local presence one of its competitive differentiators; since the acquisition, it has taken advantage of SoftLayer’s relatively small-scale “pod” architecture to expand the service from three countries to 16. The eventual rollout of its NGI architecture is likely to help IBM evolve beyond its current status as a hosting-scale provider, making it more viable for IBM to match the cost economics of the market leaders.
- The current offering is SoftLayer infrastructure, not NGI. Other than an early 2015 introduction of new storage options, SoftLayer’s feature set has not improved significantly since the IBM acquisition in mid-2013; it is SMB-centric, hosting-oriented and missing many cloud IaaS capabilities required by midmarket and enterprise customers. The details of the future NGI-based cloud IaaS offerings have not been announced. IBM has, throughout its history in the cloud IaaS business, repeatedly encountered engineering challenges that have negatively impacted its time to market. It has discontinued a previous attempt at a new cloud IaaS offering, an OpenStack-based infrastructure that was offered via the Bluemix portal in a 2016 beta. Customers must thus absorb the risk of an uncertain roadmap. This uncertainty also impacts partners, and therefore the potential ecosystem.
- The IBM Cloud experience is currently disjointed. Some compute capabilities, such as the IBM Bluemix Container Service and OpenWhisk, reside in Bluemix, but Bluemix is hosted in just three SoftLayer data centers, and is thus not local to most SoftLayer infrastructure. Some SoftLayer infrastructure can be provisioned through the Bluemix portal, but this is not currently an integrated IaaS+PaaS offering, because Bluemix and SoftLayer do not share a single self-service portal and catalog with a consistent CLI and API; do not provide customers with a single integrated low-latency network context; and do not offer a unified security context that allows the customer self-service visibility and control across the entire environment. Customers can use their IBM ID to sign in to either portal, but using Bluemix capabilities in conjunction with SoftLayer infrastructure is otherwise similar to using Bluemix in conjunction with any other third-party provider’s infrastructure.
Microsoft is a large and diversified technology vendor that is increasingly focused on delivering its software capabilities via cloud services. Its Azure business was initially strictly PaaS, but Microsoft entered the cloud IaaS market with the launch of Azure Virtual Machines in June 2012 (with general availability in April 2013).
Offerings: Microsoft Azure offers Hyper-V-virtualized multitenant compute (Virtual Machines), with multitenant storage, along with many additional IaaS and PaaS capabilities, including object storage (Blob Storage), a CDN, a Docker-based container service (Azure Container Service), a batch computing service (Azure Batch) and event-driven “serverless computing” (Azure Functions). The Azure Marketplace offers third-party software and services. Enterprise-grade support is extra. It has a multi-fault-domain SLA. Colocation needs are met via partner exchanges (Azure ExpressRoute).
Locations: Microsoft calls Azure data center locations “regions.” There are multiple Azure regions in the U.S., Canada, the U.K., Germany, Australia, India, Japan and Korea, as well as regions in Ireland, the Netherlands, Hong Kong, Singapore and Brazil. There are also six regions for the U.S. federal government; two are dedicated to the Department of Defense. (The two Azure China regions are part of a separate service operated by 21Vianet Group.) Microsoft has global sales. Documentation is available in English, French, German, Italian, Spanish, Portuguese, Japanese, Korean and Mandarin. Support and the service portal are available in those languages, plus Czech, Dutch, Hungarian, Polish, Russian, Swedish and Turkish.
Provider maturity: Tier 1. Microsoft’s strong commitment to cloud services has been rewarded with significant market success.
Recommended mode: Microsoft Azure appeals to both Mode 1 and Mode 2 customers, but for different reasons. Mode 1 customers tend to value the ability to use Azure to extend their infrastructure-oriented Microsoft relationship and investment in Microsoft technologies. Mode 2 customers tend to value Azure’s ability to integrate with Microsoft’s application development tools and technologies, or are interested in integrated specialized PaaS capabilities, such as the Azure Data Lake, Azure Machine Learning or the Azure IoT Suite.
Recommended uses: General business applications and development environments, especially those that use Microsoft technologies; migration of virtualized workloads for Microsoft-centric organizations; cloud-native applications (including Internet of Things applications); and batch computing.
- Microsoft Azure is second in market share, not only in IaaS, but also in integrated IaaS+PaaS. It has sustained a very high growth rate over multiple years, and Gartner estimates its end-of-2016 revenue run rate as approximately $3 billion. Azure is already a very capable and broad platform, and Microsoft continues to accelerate its new-feature velocity. Microsoft is now launching innovative Azure capabilities of its own, rather than primarily copying competitor capabilities. Microsoft is leveraging its tremendous sales reach and ability to bundle Azure with other Microsoft products and services in order to drive adoption. It is steadily growing the size of Azure customers; many are beginning to spend more than $500,000 a year, and a few exceed $5 million in annual spending.
- Microsoft is frequently chosen as a strategic cloud provider by customers that are committed to Microsoft technologies or that like Microsoft’s overall cloud strategy, which spans IaaS, PaaS, SaaS and on-premises solutions. Furthermore, many customers that are pursuing a multicloud strategy will use Azure for some of their workloads, and Microsoft’s on-premises Azure Stack software may potentially attract customers seeking hybrid solutions. Microsoft’s increased openness — including immediately supporting Linux in new Azure feature releases (rather than initially supporting Windows only), embracing open-source technologies and working collaboratively with a range of partners in areas of technology innovation — represents a vital and positive strategic shift.
- While Microsoft Azure is an enterprise-ready platform, Gartner clients report that the service experience feels less enterprise-ready than they expected, given Microsoft’s long history as an enterprise vendor. Customers cite issues with technical support, documentation, training and breadth of the ISV partner ecosystem. Microsoft is actively addressing these issues and has made significant improvements over the last year. However, the disorganized and inexperienced ecosystem of managed and professional service partners makes it challenging for customers to obtain expertise and mitigate risks, resulting in greater reluctance to deploy production applications or conduct data center migrations. Azure Fast Start implementations by Microsoft professional services are inconsistent in quality, and do not always accurately reflect what a customer will need to deploy production applications in Azure.
- While Microsoft continues to steadily improve capabilities that help Azure fulfill enterprise needs for security, availability, performance, networking flexibility and user management, not all such functionality is currently implemented with the level of completeness, ease of use or API enablement desired by enterprise customers. Multiple generations of solutions, coupled with unclear guidance on when to use each, create significant complexity in determining the right implementation. Most Azure customers use the portal for manual management, rather than taking a more automated or DevOps approach. DevOps-oriented customers may encounter frustrations with a lack of strong Azure support in some open-source and other third-party tools and software.
Oracle is a large, diversified technology company with a range of cloud-related products and services. In late 2015, it launched its first public cloud IaaS offering, the Oracle Compute Cloud Service (“Gen 1 Cloud”). In November 2016, it launched its next-generation offering, Oracle Bare Metal Cloud Services (BMC Service, or “Gen 2 Cloud”). In 2016, Oracle purchased Ravello (a cloud service that runs as an overlay on top of third-party clouds as well as Oracle’s IaaS), and in 2017, Dyn (a managed DNS provider); neither is in scope for this Magic Quadrant, but are closely related businesses.
Offerings: The Gen 2 service offers both paid-by-the-hour, KVM-virtualized VMs as well as bare-metal servers (including a one-click installation and configuration of Oracle Database, RAC and Exadata) and a Docker-based container service (Oracle Container Cloud Service). The Gen 1 service offers paid-by-the-hour, Xen-virtualized VMs. Oracle Cloud Machine provides a Gen 1-compatible, on-premises private cloud IaaS offering. Oracle also offers two forms of object storage (Oracle Storage Cloud Service and Oracle BMC Object Storage Service).
Locations: The Gen 2 data centers are grouped into regions, each of which contains at least two availability domains (data centers); there is a western U.S. region and a U.S. East Coast region. The Gen 1 data centers are located in the central and eastern U.S., the U.K., and the Netherlands. Oracle has global sales. The Gen 2 service is available only in English. The Gen 1 service is supported and documented only in English, but the service portal is also available in French, German, Italian, Spanish, Russian, Portuguese, Japanese, Korean and Mandarin.
Provider maturity: Tier 2A. Oracle’s cloud IaaS strategy has evolved over time. It has made several previous forays into the market, and it recently introduced a new platform (Gen 2).
Recommended mode: The Gen 2 service will appeal to both Mode 1 and Mode 2 customers, especially those with performance needs that are well-suited to bare-metal servers, and those that do not need more than very basic cloud IaaS capabilities. The Gen 1 service will appeal to Mode 1 customers.
Recommended uses for Gen 2: Cloud-native applications or batch computing that requires bare-metal servers, Oracle Databases or other use cases that require bare-metal servers to be provisioned within minutes.
Recommended uses for Gen 1: General business applications or development environments that require basic cloud IaaS capabilities, for customers that are strategically committed to using Oracle solutions.
- Oracle’s Gen 2 offering, despite being branded “Bare Metal Cloud Services,” is really both a bare-metal and virtualized cloud IaaS platform. Oracle intends for this platform to be the basis for its future PaaS and SaaS offerings as well. It is being built by a highly experienced engineering team recruited primarily from hyperscale cloud providers. It has well-designed hyperscale cloud architecture, and a thoughtful selection of current and future features. 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.
- Oracle’s broader cloud strategy spans IaaS, PaaS and SaaS. It has a strong developer ecosystem as well as a vital anchor — the Oracle Database. Regardless of whether or not Oracle is able to compete successfully for stand-alone cloud IaaS business, its Gen 2 offering will be important for Oracle’s PaaS and SaaS businesses. Customers that need to run Oracle RDBMS configurations that require bare metal, but want to do so in a cloud business model, may also be attracted to this offering, even if their infrastructure primarily resides in another cloud provider, once low-latency connectivity to other cloud providers becomes available.
- The Gen 2 offering is currently a bare-bones “minimum viable product.” It contains only the most vitally necessary cloud IaaS compute, storage and networking capabilities. It has a limited operational track record. Most customers are dependent upon direct support from Oracle’s engineering team. Oracle has just begun to build a partner ecosystem. Customers need to have a very high tolerance for risk, along with strong technical acumen. Nevertheless, Oracle field sales and executives are very aggressively promoting the Gen 2 offering. Customers should be cautious of high-pressure sales tactics, understand the reality behind the marketing and not feel obliged to evaluate the offering at this stage of its maturity. Gartner strongly encourages prospective customers to speak with references.
- The Gen 1 offering is a basic cloud IaaS offering with little in the way of differentiation, and is primarily purchased as a base for Oracle’s PaaS offerings. However, it is consistent with Oracle’s Cloud Machine private cloud IaaS offering, and thus may be attractive to customers that are interested in a hybrid cloud solution — but such customers also will be aligned to the legacy Gen 1. Nevertheless, the Gen 2 offering will be Oracle’s primary cloud IaaS offering going forward, and Gen 1 customers should factor this into their future planning.