Why Enterprises Need to Unify ML, Analytics, and Cloud
Times are changing, and the traditional models of analytics and data management don’t serve the needs of the modern enterprise, so the way to address these topics is changing too. While organizations are moving more workloads to the cloud, many mission-critical workloads remain on-prem. End users, data stewards, governance groups, and security groups alike can easily get overwhelmed with multiple access points, inconsistent user interfaces, and overall complexity. This can lead to significant productivity challenges for end users, and possibly even security breaches, too.
The typical approach has been on-premises data centers, stuffed with racks of dedicated hardware for single-purpose applications, available to only a few people. To do anything new required submitting requests to the IT infrastructure and operations teams, and an endless wait for their response. There were many, many limited-use silos of these rigid technology stacks, serviced by proprietary vendors who could dictate their prices, and the best you could say for them was that they were mature.
Today, people have higher expectations. They want the freedom to explore their business data and understand what it means. They want to discover all relevant information, not a subset. They want fast and easy self-service. They want to bring a rich variety of tools to bear, whatever fits the need of the moment, be it simple query or search or advanced data science or real-time streaming analytics. If it is too hard to do what they need to do, they’ll abandon the task, and nothing is learned. Cloud services promise resources and apps on demand.
Meanwhile, IT hasn’t abandoned the enterprise requirements for security and governance of data. IT still needs to deliver high performance at extreme scale and to do so very, very efficiently and reliably. IT doesn’t want to be the gatekeeper for users, nor expend so much effort managing the drudgery of hardware and data requests. Much better to focus on higher value, differentiated strategy. Here too, cloud promises to simplify.
A common surprise is that some cloud services don’t operate well as an integrated platform. Cloud offers solid and elastic infrastructure, but actually complicates the integration of analytics services. Most often, cloud ends up recreating the application silos of the past, only more so, because of the easy way anyone can upload a dataset and spin up a new application. This introduces complexity and risk, and increases cost substantially. Worse, many cloud analytics services are still too immature and non-standard to be readily controlled and are lacking the enterprise management tools that are required.
What is needed is a way to bring the best of both worlds together and meet everyone’s needs. Cloudera Enterprise 6.0 delivers the modern platform for machine learning and analytics, optimized for the cloud. Each of the words in the last sentence is important.
Our platform is modern – meaning it is expansive and open and flexible and capable and enterprise-ready. It makes machine learning accessible and collaborative and easy to put into production. It enables a rich data warehouse experience, only with more fluidity and exploration of ad hoc questions. And can operate smoothly everywhere – in your data center or in your choice of cloud infrastructure.
With Cloudera Enterprise 6.0, there are new possibilities for finding valuable analytics insights. In the realm of machine learning, for example, data scientists can now accelerate deep learning by 5x-10x by utilizing specialized resources like GPUs. Apache Kafka and Apache Spark are now standard parts of our platform distribution, and along with automated jobs scheduling, these make it easier for these teams to efficiently build their data pipelines and apply streaming analytics.
We’re seriously upgrading our search capabilities with Apache Solr 7 to provided integrated search alongside our Analytic Database and your business analysts’ preferred choice of BI tools. This helps the enterprise build a modern data warehouse that is far more versatile than traditional approaches.
Cloudera SDX delivers security, governance and information lifecycle management for ever more data that spans on-premises, multiple cloud platforms, and hybrid cloud. Built on a unified data catalog that contains glossary definitions, curated classifications for both discovery and governance, and end-user crowdsourced classifications, Cloudera SDX provides easy access for users and universal stewardship and controls for the enterprise.
The announcement of Cloudera Enterprise 6.0 is the beginning of a new era. The upcoming release will showcase an update of all the diverse components of our platform. We’re bringing efficiency, innovation, and enterprise-quality, and making it work for everyone. Check out the release and join us as we move into the future of data and analytics.