Google Cloud’s ambitions are far-reaching
Tuesday, August 8 2018
Google Cloud made several announcements at its recent Next 2018 conference in San Francisco, unveiling a plethora of major and minor releases. This report highlights the key points that both competitors and partners should know about.
Winning in the cloud requires an on-premises strategy as the world moves to hybrid
The major cloud providers are now launching on-premises solutions, positioning themselves to have a greater influence on the entire infrastructure market. The latest example is GKE On-Prem (Google Kubernetes Engine On-Premises). Enterprises can now run a validated GKE environment in their own data centers and register it in their Google Cloud Console, allowing them to run both environments from a single management pane. Unified management is a strong proposition, but Google also wants more enterprises designing and running their applications in Kubernetes environments, at a pace at which they are comfortable, giving them greater choice in the deployment of those applications. Taking the infrastructure “location” out of the equation is one less barrier to businesses developing and deploying cloud-native apps.
The overarching goal, and the real battle to be won, is to capture the developer team departments within enterprise organizations. GKE On-Prem is just one announcement under the broader launch of the Google Cloud Services Platform, which itself is a series of offerings and solutions to aid in the development and running of cloud applications. Istio (microservices management), Knative (Kubernetes middleware), Apigee (API management) and Stackdriver (infrastructure monitoring) are just some of the products offered in the Google Cloud portfolio, designed to make Google Cloud an attractive home for applications.
Google is forging ahead in artificial intelligence
One of Google’s longer-term ambitions is to become the outright leader in artificial intelligence (AI). As part of this aim, Google developed its own custom ASICs, which it calls Tensor Processing Units (TPUs), designed to work with Google’s TensorFlow machine learning software. Google took another step forward in its AI journey with the announcement of its first range of Edge TPUs. Effectively, these lightweight versions of its TPUs are designed to perform “inference”, which is the act of carrying out a “learned” task. One example of a business use case comes from an early adopter in LG that uses Edge TPUs in its production line inspection devices to spot faults in TV screens.
Google is positioning itself as the end-to-end platform for AI, where the compute and data-intensive aspects of machine learning occur in Google Cloud running on these TPUs, while devices with imbedded Edge TPUs execute tasks. These chips are not designed to replace CPUs but to run specific tasks in the growing intelligent connected device market.
Having dominated Internet search, Google has access to troves of user data vital to AI. The next challenge it faces in this journey is building and supporting an ecosystem of OEMs using Edge TPUs in their devices, which could become a mammoth undertaking. Producing silicon for internal use allows Google to operate at its own pace. But selling TPUs into a hardware ecosystem requires established processes in supply chain, technology roadmaps, development support kits, servicing, procurement, etc. These problems will increase at scale, as more customers look to deploy these technologies on more devices.
G Suite needs to compete on its own merits
G Suite, Google’s answer to Microsoft Office, is another important part of the Google Cloud portfolio. Google claims adoption among business customers continues to grow, in addition to its established customer bases in the consumer and (US) education markets. It announced several feature roll-outs for G Suite. For example, new security and data protection features, such as the ability to manage the regional location of data at rest, will help improve the confidence of business leaders still wary of cloud-based application services. Integration with enterprise collaboration vendors, such as Cisco, will help tap into larger installed bases. Further enhancements, such as machine learning-based grammar checking, highlight the ongoing innovation in the portfolio, but in themselves these features are not going to drive customers to switch to Google. One notable announcement was the release of Drive Enterprise as a standalone SKU. This allows Google Drive (its cloud-based storage and file-sharing service for businesses) to compete in the business sector as a standalone offering when customers are not ready to fully adopt the G Suite portfolio. This threatens Box and Dropbox, both of which have comparatively narrow portfolios and will be under pressure from investors to drive user growth.
Google is fighting on multiple cloud fronts. G Suite taking on Office 365 is a very different battle to Google Cloud Platform against Microsoft Azure. Many Google Cloud Platform customers remain Office customers, as overhauling end-user-facing applications is a separate and significant operation that takes a lot of energy to implement. Google may hold a long-term bet that winning in the education sector will drive the next generation of workers, but that is a slow tactic that gives ample time for Microsoft to retaliate. Already, Microsoft has made significant strides from a features standpoint in bringing real-time collaboration to its Office 365 products – a previous differentiator for G Suite. Google needs an army of G Suite partners to help drive organizational change; those that can resell, migrate, manage and teach customers to take advantage of the G Suite portfolio. Without these go-to-market initiatives and investments, G Suite will remain only a challenger to Office’s crown.
For more information:
Canalys APAC: Singapore Tel: +65 6671 9399 or
Shanghai, China Tel: +86 21 2225 2888
Canalys Americas: Palo Alto, USA Tel: +1 650 681 4488
Canalys EMEA: Reading, UK Tel: +44 118 984 0520