when relevant content is
added and updated.
The recent Nordic Cloud Summit in Stockholm presented Google as a new, enterprise focused cloud and infrastructure company. Today, Google is still a minnow in the global enterprise cloud market. But that will change, according to the Google speakers led by Eva Fors, Head of Google Cloud in the Nordics.
Google provided no market figures, but management claims that GCP is the fastest growing part of Google’s business. However, Nordic system integrators like Tieto that support cloud migration, put the revenue accruing from the Google Cloud Platform (GCP) in the 1-2% range of total cloud migration revenues. Google is targeting global-3000 companies in the less regulated, data intensive vertical markets like retail and manufacturing. GCP is already beefing up technical support for its indirect sales channel comprising 10,000+ ISV partners.
Google is clearly determined to commercially exploit its PaaS potential, especially after Diane Greene took over the Google cloud business in 2015. This is what the Google global network looks like on a slide from the Summit:
GCP operates across Google’s global infrastructure. Over the past six years, Google claims to have invested more than 30 billion USD in its global infrastructure. Today, 100,000 miles of fibre and 8 subsea cables support Google’s own traffic. Google claims that its network is traversed by 40% of global Internet traffic. The infrastructure has more than 100 edge points of presence and more than 800 ‘global cache edge nodes’.
Google GCP vs. Amazon AWS?
GCP’s direct competitor is Amazon Web Services (AWS) with its Migration Acceleration Program (MAP), designed to help enterprises migrate existing workloads to AWS. MAP provides consulting support, training and service credits to mitigate risk, build a strong operational foundation and help offset the initial cost of migrations. It includes a migration methodology for executing legacy migrations in a methodical way as well as robust set of tools to automate and accelerate common migration scenarios.
The Google GCP strategy seeks to avoid directly competing with AWS pricing and global availability. Instead the focus is on a wide range of ancillary apps and services that go far beyond what AWS offers. Essentially GCP rests on three basic pillars.
The three GCP pillars
- Security: For enterprises to reap all the benefits of cloud computing they need to trust cloud data centres and the Internet connections to them as much as/more than? they trust their own in-house data centres and corporate networks. Since the emergence of public cloud services, their security record has been a lot better than most corporate data centres, and generally CIOs have come to respect the security provided by tier 1 cloud providers. Google wants to demonstrate the same levels of security for data across its own infrastructure.
- Google supports its customers in preparation for GDPR (General Data Protection Regulation), especially with regards to data centre management, data sovereignty and PII protection (Personally Identifiable Information). One tool demonstrated at the Summit is the DLP API, a layered security service which redacts PII data on retained documents or chat conversations. Typically, this includes contact information, faces, data on credit cards, ID document data etc.
- Transition and efficiency tools: The shift from on-site computing to cloud computing involves very different steps depending on industry vertical, geo locations, size and maturity of the company making the move etc.; a lot of support tools are needed. Google wants to make that journey with the customers and develop tools to facilitate both transition and operations. Google claims to have developed over 500 of such tools just in the past 6 months.
- The cloud based G Suite facilitates the creation, connection, access and control of corporate communications, and automating and reducing time spent on coordination activities.
- Analytics: Google is synonymous with big data, and data crunching using. It’s all about a company’s ability to collect, mine and extract useful information fast from vast data hoards. Today, many data scientists are bogged down with maintenance tasks e.g. maintaining a Hadoop Platform. This is not a good use of their time. Any loud platform, including GDP, has the potential to take that pain away.
GCP analytical tools
- At the Summit Google demonstrated several analytics tools, including:
- BigQuery: Google’s serverless fully managed, petabyte scale, enterprise data warehouse for analytics.
- G-Sheets: Addressing spreadsheet complexity with natural language input and the ‘Explore’ button with a range of data analysis and graphic presentation
- Cloud Spanner: Removes the schism between SQL and No-SQL DBs with simple querying and scalability options
- TensorFlow: an open source software library for building machine learning systems using data flow graphs.
- Cloud Machine: machine learning engine that delivers a managed TensorFlow service
- Vision API: Detecting labels, faces, colours
- Video translation API: understands your video
- Speech API: speech in, text back
- Natural language API: analysing and routing feedback from customers
- Translation API: Google Translate now uses neural network translation to create a higher layer language to translate between languages and words not previously compared.
- Hardware acceleration with Tensor Processing Unit. This is a custom ASIC that will become commercially available this year.
Google GCP vs. Hybrid Cloud?
Google’s strategy is to be a one stop shop for all corporate apps and infrastructure. Hybrid cloud is merely an intermediate step on the way to total cloud computing, where the real performances advantages accrue. Many analysts would disagree with this notion, but actual analyst data supporting the hybrid cloud adoption is inconclusive.
Recent Quocirca research on hybrid cloud adoption indicates that overall perceptions of cloud computing are reasonably good. Expectations are being met in most areas. However, when it comes to implementing and using a hybrid cloud, there are issues. They center on technical and human security, as well as data sovereignty, costs and performance. Areas stated as being catalysts for organisations to more rapidly embrace cloud include better overall support for standards, the use of automated APIs and automated workload management across the total logical platform. These are all points that GCP addresses.
So Google may have a point. The pains of hybrid cloud interaction and management may be alleviated by putting all the heavy data driven apps on the GCP cloud, and merely retaining financial and admin systems on-site.