Todas las entradas hechas por johnarrasjid

The cloud goes critical in 2018: Deep learning, smart cloud infrastructure, and more

From cut-throat competition, eyebrow-raising co-opetition, and major advances in cloud-based machine learning, 2017 was a pivotal – and productive – year for the cloud, setting the stage for what looks likely to be the most exciting year yet. 

The market swing is already in full force. Thanks to a full-fledged embrace by the enterprise, the cloud is undergoing dramatic transformation as vendors rush to meet the infrastructure and business needs of today's top companies. According to Gartner, the overall market likely grew by close to 20 percent in 2017, and IaaS in particular saw close to 40 percent growth. With digital transformation at the top of every executive's mind, it's likely that this trend will only accelerate. In fact, by 2020, Gartner estimates that the overall market will reach a whopping $411 billion, and IaaS $72 billion, 87 percent and 185 percent raises respectively from 2016. 

What we considered in the realms of "crazy" 10 years ago is now a reality, and the leapfrogging will continue. Based on our collective experience, interactions with customers and conversations with colleagues, here are four key trends we see unfolding in 2018: 

Deep learning 

There's no doubt that 2017 was the year of machine learning. Culminating in the blockbuster announcement of Gluon, a brand new cloud-based open source machine learning platform, as well as other advances in technologies, ML is finally set to become a real part of the enterprise business strategy. 

While advances in ML will continue in 2018, expect this to lead major breakthroughs with deep learning as well. According to a new survey from Vanson Bourne, 80 percent of enterprises already have a form of AI in production today, and 30 percent are planning to expand their capabilities over the next three years. Cloud providers are already anticipating this need, as some already allow enterprises to leverage GPUs (the key piece of technology for deep learning) for massive parallel computational power. Expect to see an explosion of deep learning, as the costs for this service drop thanks to commodification and more cloud providers offer the service. 

Smart cloud infrastructure 

As major advances in automation and machine learning continue to make pace, expect to see the beginnings of more smart and automated cloud infrastructures, ones that go beyond traditional automation and can actually make seemingly human-like decisions about important issues around authorisation, security, vMotion, dynamic resource scheduling, load balancing, and self-healing environments. This will change the way IT departments approach technology, with the same impact that virtualisation had when it was introduced to the market in 2003. 

The "instant" private cloud 

There may be quite a few disadvantages for enterprises using a public cloud, but there is one main draw that keeps people coming back: the simplicity of setup. As a result, despite all the drawbacks, many companies continue to leverage public cloud for this reason alone, but with an eye to providers focused on mission-critical workloads. 

In the coming year, expect mission-critical cloud providers to bring the public cloud experience on-premises. Private cloud vendors will provide their own "one-click" setups for customers that will go well beyond just having a server that is instantly up and running. Instead, you'll get one that is pre-configured to the specific needs of your enterprise before you even turn it on. 

Continued co-opetition among cloud vendors 

Although 2017 had its fair share of drama, it was also very much the year of market maturation. For example, VMware teamed up with AWS and Pivotal announced a partnership with Google Cloud, all for the good of the customers. 

Cloud providers have three options: the expensive route – they can go it alone, spend hundreds of millions of dollars on acquisitions, or they can partner with their competitors to build ecosystems that meet the unique needs of their customer base. 

This year, expect to see more interesting (and unexpected) partnerships develop, as cloud providers compete to meet the needs of customers. It certainly looks as if 2018 will be the year of the specialised cloud.