Edit | Amao Design | Mu You
For thousands of years, the advancement of technology has always been closely linked with the development of human society, and every major technological change has had a profound and lasting impact on society. The symptoms of the past half century are especially obvious.
Digital technology is nourishing all things with its attitude of "spring breeze turning into rain", bringing unprecedented great improvement to human society. People's living habits, working methods, and even social cultural customs, health and well-being, etc. have also "changed their appearance over the years."
Every time at the end of the year and the beginning of the year, even ordinary people who have nothing to do with information technology also hope to hear about the trends and trends of technological development, so as to open a window into the digital world and make new insights into future work and life. Visions and visions.
John Roese, Global Chief Technology Officer of Dell Technologies , may be one of the best people on the planet to interpret and answer digital technology. After all, before assuming his current position, he had served as the CTO of well-known companies such as EMC, Futurewei, Nortel Networks, Broadcom, Enterasys, and Cabletron System.
For a long time, John has been concerned about and involved in all aspects of digital technology, such as software, hardware, network, communication, storage, computing... His comprehensive vision is breathtaking. At a recent communication meeting exclusively for Chinese media, John shared his observations and thoughts without reservation.
Or, more accurately, advice.
Unexpectedly, when more and more enterprises are entering the stage of using the cloud for their core business systems and using the cloud in depth, John's first suggestion is "Without knowing the long-term cost, I will not use the cloud" .
Does it mean that it deviates from the current "mainstream"?
In fact, John expressed the true aspirations of many companies. In the past few years, many enterprise customers, especially those who have adopted public cloud infrastructure, have often been surprised by budget overruns.
In the service-oriented environment of the public cloud, enterprises are actually faced with various variable costs, which makes it impossible for the IT department to accurately grasp them.
The crux of the matter is that IT departments typically allocate a fixed budget in advance, so when a situation arises that causes budget overruns, they may have to make some unnecessary trade-offs, or even postpone other important projects.
John cited a project where a customer implemented AI+machine learning as an example. Such projects, he says, typically break down into three distinct missions:
The first is to develop AI algorithms. At present, the public cloud is the fastest, cheapest, and best place to develop AI algorithms. There are very good tools in it, and the start is simple, and the initial algorithm development does not need to cost too much.
The second is the training of the model. Similarly, the public cloud is also an excellent place to train the AI architecture. Due to its good elasticity and huge scale, we can conduct various training on the model and turn it off when not in use, which can well control the effect and cost.
The last is the application of AI algorithm. Although the public cloud is used in the previous two links, more and more customers find that there is no suitable application environment for the public cloud in the real world, no matter factories, hospitals, or other links of transportation. In other words, the final application environment of AI algorithms should not be the public cloud in most cases, but should be placed on the edge.
In this way, John believes that when customers go to the cloud, they should not only consider technology and reputation, but also fully consider the short-term and long-term costs of using the cloud environment, and be aware of it, so that there will be no budget overrun caused by going to the cloud.
John's second suggestion is related to zero trust security .
In view of the current new IT technology architecture, zero trust breaks the borderline network defense idea, and emphasizes the continuous verification of any person, event, or thing trying to access the network, which has been proved to be a concept more suitable for modern application architecture.
John said that zero trust is not a new concept, but as of now it is still not widely implemented. In his view, Zero Trust represents three paradigm shifts, bringing about three massive changes in security.
First, one of the important pillars of zero trust is that various known and unknown devices, events or users in the architecture need to be verified in the application and must be continuously authorized, which is also a complete violation of traditional security. subversion.
Second, in terms of strategy, traditional security protection is to define bad things and keep them out. Zero trust is different. It will describe the known "good things" in detail, and nothing else will be allowed to appear, that is, only good things will be allowed to happen-to prevent them from happening.
Third, previous threat management and detection tools are often outside the infrastructure, so it is difficult for them to gain insight into threats. Under the concept of zero trust, threat management and detection systems are embedded in the infrastructure, so potential anomalies and threats can be dealt with in a more real-time manner.
Zero trust has brought a huge change to infrastructure security, and it has also been verified in all directions. However, so far, many organizations and institutions have not taken action on this, many of which are difficult to implement.
Dell has previously announced initiatives to accelerate customer adoption of zero trust, John said. Through a series of measures such as consulting and services, Dell hopes to help organizations and even the entire ecosystem, strengthen and simplify existing product portfolios, and even deliver reference architectures...all of which will continue to develop and deepen in the next few years .
“The first step in the zero trust journey is to define your control plane, identity and maintenance management tools. Even without zero trust, (enterprises) have a clear security control plane, which is extremely high in the current traditional security environment. the value of."
John's third suggestion is about quantum computing . But perhaps for some reason, he is willing to split it into two parts, 3a: building early skills to take advantage of quantum computing, and 3b: determining where to deploy quantum-safe encryption first.
We have obviously not reached the stage of encryption in the form of quantum systems today, which means that in the future, in the presence of a large number of quantum computers, our current work and algorithms will be broken and invaded like autumn leaves .
"The encryption technology that is still very secure today will be difficult to prevent the decryption of quantum computers in the future. Although we don't know when this day will come, it must be inevitable." John pointed out.
So, how should this risk be dealt with?
Technically, there should be a post-quantum cryptography. At present, Europe and the United States have made corresponding preparations for it, and I believe China will also define it. Using new algorithms and encrypted defense lines or keys to manage data has become a recognized choice in the industry.
The problem is that many customers may not know exactly where they should encrypt? John also gave three steps:
1. Inventory resources, find out which links in the data center and cloud computing interface of the entire enterprise resources need to be encrypted; 2. Inventory the encrypted resources, especially the link of data exchange between private cloud and public cloud; 3. When the quantum algorithm becomes a reality, deploy the quantum algorithm on the back end of the public interface facing the outside.
"2022 is a very active year for quantum innovation." John said that in all walks of life, people have begun to think about how to use quantum computing-not in the distant future, but in the foreseeable few years. Research and development of batteries, etc.
He suggested that enterprises should carry out quantum research as soon as possible and establish corresponding skills preparations, so that when quantum computing becomes a reality, enterprises can benefit in the first place. "If a company's CIO does not understand the relationship between quantum computing and its own business, the company has actually fallen behind."
Quantum computing is not out of reach, and some companies, including Dell, have launched corresponding systems. With these solutions, companies can also enter the world of quantum computing without owning a quantum computer.
John's fourth suggestion, around the edges - this one is very important.
He believes that the vast majority of customers do not have an edge architecture for a multi-cloud environment, but only a cloud expansion model. Therefore, if a certain public cloud service is used, the public cloud will be extended from the edge to the edge.
So, is there any trouble doing this?
The answer is yes. Typically, an enterprise will use 3 to 4 public clouds, so if an edge environment is created for each public cloud, this kind of edge proliferation is inevitable.
What's even worse is that the above model is expensive, ineffective, rigid and inflexible. However, this is what most companies are still doing. Some leading manufacturers, including Dell, hope to effectively change this situation.
Almost a month ago, Dell announced Project Frontier , a multi-cloud "edge platform" on which enterprise customers can support any software-defined edge, whether it comes from the cloud, Internet of Things, or other environments.
"Whether in a factory or a hospital, Project Frontier can dynamically deploy different cloud edges in the environment. Customers can deploy it as a software function, and just uninstall it when it is not needed," John said.
It sounds really cool, allowing the edge to be plug-and-play, unplug and stop, which is also an application feature that the majority of enterprise customers have been looking forward to for a long time. It is reported that Project Frontier will be delivered from next year, and I believe this will be one of the most interesting and important contents in the enterprise field in 2023.
Previously, enterprise customers often had difficulty making decisions when faced with multi-cloud architectures, and edge platforms like Project Frontier will help them make the most accurate choices based on technology and application development trends.
"All CIOs should start from January 1, or as early as possible to think, make decisions and plan for the above recommendations. If CIOs really do this, it means that enterprises have seized the technological innovation in 2023 waves and trends.” John concluded.
Futurist Negroponte once said, "The best way to predict the future is to create the future." In the face of a technological future that is difficult to accurately predict and grasp, Dell will continue to use decades of profound insights and practical practices We are committed to helping more organizations and institutions achieve greater success with leading and complete product solutions.
