Google: Who owns the DNA of the next generation cloud platform? Machine learning and cloud intelligence are the future of the cloud

This is a creation in Article, where the information may have evolved or changed.

At the "NEXT2016" meeting, Google's Eric Schmidt said one of Google's biggest strengths was standing at the forefront of cloud computing for the next 10 years. It's not infrastructure or software, and it's not as simple as pure data.

crowdsourced intelligence is an evolution that can create smarter systems that users can use, retrain, change, and output for their own purposes, from top to bottom. All this will ignite the future direction of the next generation cloud platform.

This concept of crowdsourcing data makes training better, smarter, and broader, which is the core advantage of Google's competitiveness in the next 10 years of cloud computing, Schmidt said. "I firmly believe that the rapid assessment model, Google Cloud computing, machine learning, and crowdsourcing will be the basis of every IPO and be a winner in the next 5 years. "It would be a very similar assessment to the process that generated the application--and then put it in the smart creation that the user really needs right now, behind which there is the underlying infrastructure, and many of these levels of abstraction will give the user a seamless experience." Schmidt said it was very similar to the car industry-from the early days, to the clutch, to the future.

"The Google Cloud Platform is perfectly synchronized with what is going to happen in the future. The platform is not the end, but the bottom--there's always something on it. This "What" is machine learning, including the narrow sense of AI and generalized AI, is the transformation of the next generation. The programming paradigm changed so. In addition to computer programming, you can also make your computer do what you want it to do. This is a fundamental change in programming. ”

Schmidt said that the best preparation for the transition was to start with Linux, above and above the Google Cloud platform and Kubernetes. From a development perspective, it is designed to work in a scalable, portable language like Go,python,node.js,java and created on Google App engine. But on this list, it is the application of TensorFlow, which will be at the heart of Google's machine learning and analysis strategy for the next 5 years (he reckons many other businesses too).

For TensorFlow please click here

With the first display of Google's new machine learning platform, it was the first time that the banner had been inserted in the sand, followed by companies such as Aws,microsoft Azure and others with advanced machine learning and cloud infrastructure such as IBM. Because unlike other companies, Google has its own existing and many other companies do not have the services such as photos, voice and text for resources to do crowdsource services, to promote the listed services.

"In the last few years, we've plugged machine learning into everything we do," says Jeff Dean, Google System Infrastructure group, the Google Brain Project leader. "That's why the company's Android voice recognition system works so well and why their translation services are so far-reaching and relatively accurate, and that's why photo face and object recognition and classification can be seamless." The crowdsourced service is that Google provides the richest files and streaming data to make smarter machines, and more advanced deep learning capabilities. For other clouds that were previously dominated by hardware, Google would take a more modest approach (through software) to counter it.

With the installation of TensorFlow to allow complex deep learning as a comprehensive management service, the next decade of cloud computing is very different, let alone the next generation will be the world of data analysis. We have freed ourselves from the need to use the same old tools (some new frameworks like MapReduce and Hadoop) to deal with a lot of data, by letting the data compute itself to achieve a "quantum leap", at least in theory. And that would be a big deal-a big event that could threaten the current supremacy of the cloud, with Google likely to focus on the whole stack and have crowdsource-able services.

Even if tensorflow just emerged at the end of 2015, Schmidt believes this is the next generation of Google's ability to provide a deeper, broader range of services to large enterprises with complex analytic workloads. This is the beginning of the next paradigm wave of application and cloud applications – and it is likely that the top 500 companies and AWS will take their first step toward Google's own role as a driving force and the top 500 companies.

Not so long ago, as we commented in the first decade of AWS, even in very resilient applications, the focus is still on the evolution of infrastructure, and the tools and services associated with it are linked to each other. The first decade of the cloud is strongly rooted in these considerations: Consider the security, the network, and of course the need for large hosts to drive an exploding user application. It's no secret that AWS is at the forefront of the cloud platform, which is well known, as additional elements start renting backup infrastructure (compared to just creating a data center for hosting cloud-based services), as well as other companies that lack the initial motivation.

All of AWS's previous efforts deserve recognition, but now AWS and his peers in cloud computing are installing crowdsourced data, machine learning, and deep learning frameworks to do high-level business and support new existing services; Google's positioning is to allow users to get the value of service. This annual meeting is the foothold of Google Cloud to some extent, it has not worked with large-scale enterprises before, these enterprises through Amazon's services to the public cloud (usually through the hybrid model) have a preliminary attempt, need some process to be convinced to get rid of the initial experience of the fear, cost, risk, From pure cloud infrastructure to experience another shift.

The cloud platform that can make everyone's attention is definitely no longer a hardware facility. That's a pretty similar process, and it's expected that Google will make further efforts to support "instances" of cloud services that Google can provide, just as AWS has done before, and AWS is providing services in many areas of storage, acceleration, CPU, and networking options to seize the market. The core competencies must come from a thing that is hard for most companies-the next generation of intelligent analytics can be so smart and coordinated that it's a matter of disturbing the original cloud-ecological layout and changing the rules of the game in the future, at least in Google's eyes.

Original link

(if you need to reprint, please contact us oh, respect for intellectual property is everyone's responsibility)