A few days ago, Amazon’s AWS re:Invent 2016 conference just came to an end. At this five-day conference, a large wave of new products unveiled, indicating that AWS's capabilities have risen to a new level. From the current point of view, AWS accounts for more than half of the public cloud market, and its revenue is very gratifying - AWS's revenue in the third quarter of this year is as high as 3.2 billion US dollars, and the annual revenue is expected to reach 10 billion US dollars. In this regard, AWS CEO Andy Jassy said that AWS will continue to be the "important and powerful leader" in the field of cloud computing.

However, this road Amazon will never go smoothly, because Google, which has been left behind by it, is already catching up. Recently, the famous technology blogger Ben Thompson wrote an article to analyze whether Google can beat Amazon in the field of cloud computing.

The following is from Stratechery, originally titled "How Google Is Challenging AWS," by Ben Thompson.

Big companies are always inevitably criticized for "missing the future" - some are lost because of their complacency with the status quo (this should be a painful approval); some, although they have seen the need to break the status quo, but However, it is always wrapped around the feet, just around the existing products, and in the end, did not push down the wall of their own piles, and finally only across the wall, watching an era whizzing past. In this regard, Microsoft has provided us with excellent negative materials. As early as 2000, Microsoft launched Windows Mobile. From this perspective, Microsoft has had a relationship with the mobile era. But in the end, it only earned a few screams at the entrance, and saw that the door was closed.

The reason why Microsoft missed the mobile era is, in the final analysis, because: First, it is too full and not willing to give up (for example, the model of profitability by license is not allowed); Second, it does not see the future after ten years. Mobile products, including smartphones, are the "sun", and its core product, Windows, is just one of the many "planets" that rotate around the "sun."

In addition, Google has had a similar failure. Ten years ago, the company launched Google Apps for Your Domain (now evolved into G Suite) with the eyes of its peers. It should be said that Google was an early step to see the future of the office suite. However, over the years, G Suite has rarely improved, only insisting on its "low price + low configurability", the result has been tepid, can only be played between small business owners and education It’s a little trouble, and it’s always hard to stick to big companies. By 2011, Microsoft's Office 365 was born. With almost uninterrupted innovation, in a few years the rookie scored more than G Suite and became the office tool of choice for many companies. Although G Suite has been catching up recently, it has introduced functions such as machine learning, but in the face of the young and energetic Office 365, there is still some powerlessness.

Thinking about it now, 2011 should be a memorable year. That year, the cloud-based Office 365 was born, Google and Microsoft not only officially avenged in the smart office field, but also began the looting of the virgin land in the cloud. Sadly, when the battle is in full swing, neither of them noticed. Outside of them, a company that couldn’t be beaten with Cloud gossip has long been ready to become a real cloud computing giant.

This company is Amazon.

Picture from Fortune

Also in 2011, Amazon launched AWS. First, Amazon took the lead and used it. Then, enterprise customers from all over the world flocked. So, what is the charm of AWS? First, AWS itself is a system of sustainable evolution. It can accomplish tasks such as large-scale upgrades and large-scale customer transfers without downtime and downtime, allowing enterprise customers to enjoy 7x24 hours of uninterrupted service without worrying about their impact. The second (and most important), what AWS provides to users is not the frameworks but the primiTIves. With these “primitives”, customers can build a sustainable, sustainable, scalable back-end system that meets their specific needs.

AWS's "primitive" (PrimiTIves)

At the beginning of the build, Amazon engineers found that instead of providing customers with a full-featured framework that lacks the ability to upgrade (primiTIves), it is better to provide primitives (primiTIves), one of the most basic computing modules. Customers can select and use these modules to suit their needs, ultimately creating an efficient, scalable, and secure system.

I have to talk about the story of primitives here.

According to the introduction of Yibang Power Network, at the beginning of 2000, Amazon came to a bottleneck period. The situation in which teams competed for infrastructure at the time was very serious. At this time, Bezos was obsessed with a book written by Steve Grand called "Creation." Grand is the developer of the video game Creatures in the 1990s. According to the description, in the game, he only needs to provide a simple computational building block called primitives to the player. It can be cultivated into various electronic intelligent creatures according to your own preferences. This book gave Bezos great inspiration. He believes that if Amazon wants to stimulate the creativity of developers, it should not guess what kind of services they need, because such guesses can only be based on past experience. Instead, Amazon should create primitives, the basic modules of computing, and then let go, giving developers the freedom to use them. Similarly, in the face of the company's current real-world problems, Amazon needs to break down its infrastructure into the smallest and simplest atomic components, making it as easy as possible for developers to access. The "primitive" model ultimately simplifies the modularization of Amazon's infrastructure, and these modules are not only available to Amazon's internal teams, but also to external developers.

Not only that, but Amazon also used this principle in its internal management reforms. Bezos split the entire company into small, highly independent and highly autonomous small teams of less than ten people (the so-called "primitives"). These teams are small and savvy, and they can feed all the staff by ordering two pizzas during overtime work, and they can play well when Amazon encounters major problems. In short, after the spin-off, the entire team seems to be a mess, but because these "primitives" are highly self-motivated, they can grow relatively flexible, and in the end, the results they bring are amazing.

Everyone should be aware of this: Amazon used the original solution to the infrastructure problem and the internal management reforms used in the development of AWS. As a result, it was a great success. In this regard, it has been asserted that AWS provides services far beyond the reach of infrastructure (such as processors, hard drives, databases, etc.) and far more than just a software or platform service. It provides a highly flexible and highly scalable "primitive" that allows companies to freely play and ultimately create a variety of what they want.

Google has never been a platform company. Often, people will divide Apple and Google into two completely different camps—the former is the leader in “product companies” and the latter is the leader in “service companies”. People have this view because their understanding of the concept of "products" is too narrow (perhaps in many people's view, the so-called products are things that are tangible and tangible in hardware). In fact, if we reasonably broaden the extension of the "product" and call the ideal solution for the end user to be called "product", we will find that Google and Apple are very similar and belong to the product company. .

After all these years, I was selling a product (picture from Reuters)

There is actually another important reason for defining Google in this way. As we all know, all products, whether smartphones or search engines, must undergo an almost unknown process of pain before being delivered to the end user. In the meantime, they must undergo many modifications, improvements and even rework to bring them to consumers. The best user experience. So we said: In the market dominated by end consumers, integrated products are very popular. Google, while offering consumer-oriented services, is 100% "integrated" -- exactly the same as Apple's iPhone.

Obviously, neither Amazon nor IT, the dominator of the IT era, offers no integrated products or services. For example, Microsoft's Win32 API. In terms of the end user experience alone, Windows is significantly less designed than rivals like Mac OS. But on the other hand, Windows' performance and scalability are so strong that many companies rely on it to develop apps, so until now, Windows is still in a dominant position. Like Amazon's AWS. All the architectures that can build back-end services are split into highly flexible basic modules. After this, AWS soon defeated Google's cloud service Google App Engine, which was launched in 2008. In this regard, some people have made such a summary: using App Engine, you must accept Google to replace many of the decisions you make; and with AWS, you can build what you want as you like.

There is a common reason for the success of Windows and AWS. Apps developed for Windows are generally difficult to be compatible with other operating systems, but because Microsoft's partners and agents have a large business network, they have turned Windows into the only choice for many businesses. Over time, a huge ecosystem based on the Windows API and centered, but not able to get out, was formed. Obviously, AWS has been following Microsoft's -- it's about building this "who is it" ecosystem.

But this ecosystem has begun to collapse. The rise of the Web has finally enabled consumers and businesses to take advantage of both Windows and its technology; likewise, the rise of browsers has finally made AWS users no longer the only choice, because: Any enterprise app is now Built for the Web, so people can log in from any device.

As a result, Google has the opportunity to fight against AWS.

In recent years, Google has practiced a path that relies on browsers to explore enterprise computing. In 2014, Google announced the launch of Kubernetes, an open source container cluster management system based on the Borg service. Because Borg can leverage a large amount of Google infrastructure, all of Google's services can take the appropriate computing functions on demand, without having to worry about any details. The key to this process is the container. Google engineers have built a standard interface (interface) that is almost lossless, so there is no need to worry about the underlying hardware or operating system.

The difference between Kubernetes and Borg is that the former can be home everywhere. It runs on AWS, runs on Azure, runs on Google Cloud Platform and on-premises infrastructure, and even runs in the user's home. For Google, although AWS has had 10 years of experience in IaaS (Infrastructure as a Service), with Kubernetes, it has the same chance to catch up. First, Google has made great strides in its infrastructure; second, the potential impact of Kubernetes and the container-based R&D philosophy will ultimately give users the freedom to choose infrastructure vendors. No wonder Kubernetes has become the fastest growing open source project in history, because it really doesn't bind people.

But how does this help Google? After all, even if Kubernetes becomes the standard for enterprise clouds, the AWS ecosystem will never let go of a corporate customer. So what Google really needs is to differentiate.

What needs to be clarified here is that the openness of the Web itself is not the root cause of Google's confidence to defeat Amazon. However, this openness creates conditions for Google to develop the best technology. At the same time, Google has the most advanced search engine. The same thing in itself does not guarantee that Google will be able to beat Amazon. However, because its search engine relies on links rather than web content, as long as the web grows, Google will grow. This advantage is not available to its competitors.

I think we can come up with a concept that deserves to be broadened - in fact, it is also the core of Aggregation Theory: when the distribution cost or switching cost drops, the user The importance of the experience will rise. That is to say, when you can access all services (whether it is news, car sharing, video or search), the best way to win is not to win the initial advantage, or to win the advantage combination. (Note: This section of the translation refers to open source China)

So, Google is challenging AWS in the enterprise cloud space, first of all to put the treasure on the open source Kubernetes. With Kubernetes, Google hopes to create a browser that is completely uncontrolled by the cloud infrastructure and then reduce conversion costs. In addition, Google will also launch a sprint to machine learning, and the results achieved in this field will also become a weapon for Google to fight against AWS.

Winning undecided image from CloudAcademy Blog

Undoubtedly, cloud services are increasingly leading machine learning. Both involve the processing of large amounts of data, and in this respect, only a small number of giants have sufficient financial resources to eventually build the infrastructure they need and hire the world's leading machine learning engineers. This means that for most companies, the size of machine learning depends first on whether their data comes from the cloud (although there are local solutions, but I think these programs will not go over time) Applicable), and secondly depends on the quality of their respective cloud providers.

This means that cloud providers have to take on even greater risks. Excellent “textbooks” for machine learning not only need to be differentiated, but also ensure sustainability, because only excellence can attract more users and collect more data. Finally, the data will help companies improve their respective Machine learning.

It can be said that Google has become the biggest threat to AWS in the cloud in terms of the data it currently has.

Google's great advantage is that it has been dealing with large amounts of data for nearly 20 years and has been developing powerful machine learning algorithms for the past few years. You know, the data is the most important, and the best evidence is that shortly after Google announced the open source TensorFlow (a machine learning blueprint) last year, I noticed that the company wanted to take the opportunity to hint to people: its data reserves are more It is huge, and its infrastructure is also superior to sustainable evolution.

In fact, Google has been unable to show us how it uses these advantages in its cloud products - just before this Thanksgiving, Google released a series of products, from which we can clearly see that these products are used Google's advantages in data:

Cloud Natural Language API: Analyze text through machine learning, which has certain availability;

Advanced version of the Cloud Translation API: It uses machine learning to dramatically improve the accuracy of translations in eight languages;

Cloud Vision API for large price cuts: Analyze images through machine learning;

New version of the Cloud Jobs API: Assigning the right jobs to employees through machine learning

All four products have been linked to another product, the Cloud Prediction API, which relies on machine learning. And it's clear that the Cloud Prediction API and the top three of the four products are born out of Google's consumer products. Among them, the cloud work API seems to be based on Google's internal tools, but also based on the large amount of data Google has obtained from the Internet. In short, Google has spent many years polishing various algorithms, hoping that these products will perform well when applied to corporate data. Of course, I hope that this advantage of Google can continue.

Of course, Google must go further. It realized this on its own, so it was announced that the top AI experts Li Feifei and Li Jia led the establishment of the Google Cloud Machine Learning Group. The team will be responsible for the developer's API for machine learning. In other words, its role is to productize Google's machine learning capabilities.

So, after a big circle, we found that Google finally touched its own tricks. In the first wave of cloud computing wars, it was indeed left behind by AWS; but later, after Kubernetes was open sourced, Google began to turn. With the successful development of open containers that do not require any supplier requirements, Google is putting more and more energy into product development. After all, as a company, changing the way to compete is always easier than changing the essence of oneself.

To be sure, Google’s success is not a sure thing. The company still faces the challenge of a new business model, while also establishing a sales and service team as soon as possible. It should be noted that in both respects, Google lags behind Amazon – after all, the latter has a large ecosystem of partners and a more powerful service.

In addition, AWS also has its own machine learning API. In this regard, IBM and Microsoft are not to be outdone. According to statistics, Microsoft is very strong in this area - not only has been researching in this field for many years, but also has a wealth of experience to productize and commercialize technology. Also, Google has consistently pursued a "consumer-focused" strategy that is likely to be an obstacle in the future. Finally, although Kubernetes is very popular, don't forget: Google has never used it.

However, Google is still a strong competitor. It is strategically sound, and (possibly) more importantly, it is now eager to create new businesses. Finally, don't forget: the long march to the cloud computing has just begun. Although Amazon seems to be far ahead, in reality, the future of cloud computing has not really come. So let's take a look at Google and see what tricks it will make to change the "future rules" that haven't yet formed. I think this process will be very interesting.

CBD 510 Cartridges

510 cartridges,510 vaporizer,CBD THC oil Cartridge,Cannabis Oil vapes,Hemp cbd Vaporizer

Shenzhen Yingyuan Technology Co.,ltd , https://www.yingyuanvape.com