Building An End-to-End Digital Infrastructure: Intel Embraces The New Era Of Smart Edge

Against the backdrop of the integration of technologies such as IoT, 5G, AI, and big data, intelligent edge computing has rapidly transitioned from concept to reality. Especially since the outbreak of the COVID-19 pandemic, an increasing number of network applications, such as video surveillance, smart retail, smart manufacturing, and smart cities, have developed rapidly, with more and more data being stored and processed at the edge. According to Gartner’s prediction, by 2025, 75% of data will be generated outside of data centers at the edge, indicating that digital transformation is moving from the cloud to networks and endpoints. However, the intelligent edge also presents new challenges for industry enterprises. How will technology trends develop? How can end-to-end digital infrastructure be built? How can related applications be implemented to empower industries such as healthcare, agriculture, smart transportation, and retail?

Integration of Intelligence and Data: Edge Computing Takes Center Stage

The Internet of Things (IoT) is not a new concept; if we consider the proposal of the “Perception China” concept in 2009 as a milestone for the introduction of IoT in China, it has been 13 years. The concept of edge computing was officially proposed in 2013, and it has also been nearly 10 years. However, as the digital economy continues to deepen, there are higher demands for networks and computing power, and the trend of mutual integration between IoT and edge computing is becoming increasingly evident, making the intelligent edge a new industry hotspot.

According to Zhang Yu, the Global Chief Technology Officer of Intel’s Video Division, the increasing emphasis on intelligent edge computing is largely due to the profound digital transformation that society is undergoing. In this process, the operations of various industries must rely on a complete set of end-to-end digital infrastructure to function normally. The edge plays an increasingly significant role as an organic component of the entire system.

“The emphasis on ‘end-to-end’ is because the services provided by this system are all end-to-end services, from data collection at the front end, to data transmission, and then to data processing, storage, and feedback at some key nodes. The entire service is a complete end-to-end process, and only such a complete process can be effectively accomplished,” Zhang Yu pointed out. In this data transmission and processing process, some data can be completed at central nodes, while others must be processed at edge nodes. This is determined by the different demands of data applications. Applications with lower real-time requirements can transmit data from the application front end to a remote cloud, where there is more storage space and relatively abundant computing power, allowing for calculations and storage at a lower cost. However, for applications such as autonomous driving and industrial internet that require high real-time performance, transmitting data to a remote location for processing is not suitable. Such computations often need to be supported by edge computing architectures like computing power networks. This is the main reason why intelligent edge computing is receiving increasing attention.

“Without the intelligence of IoT and the ubiquity of computing power networks, it is impossible to support the real-time and efficient computing needs of applications at the front end,” Zhang Yu emphasized.

For intelligent edge computing, three future development trends are emerging. First, the construction of digital infrastructure will be an integrated system of “edge-cloud” collaboration. None of the “edge-cloud” can be absent, and they play different roles depending on the application scenarios, with the edge playing a more important role in processing some data, while at times the cloud takes on a more significant role.

Secondly, the role of software in digital transformation cannot be overlooked. This is commonly referred to as “software-defined.” With the development of remote work, the demand for network flexibility is constantly increasing. This requires a network that can be flexibly deployed, not only for the core network but also for the configuration of data centers, which must become increasingly flexible. Data centers need to be built in a programmable way, with programmable Ethernet chips and programming languages becoming important supports for data center construction.

Thirdly, the penetration of artificial intelligence into edge computing is becoming increasingly widespread. Although the current wave of artificial intelligence applications began in data centers, with a large amount of data processing initially completed in the cloud, as technology develops, more and more data processing is shifting to the edge. The mutual penetration and integration of edge computing and artificial intelligence have made the trend of AI ubiquity inevitable.

In summary, against the backdrop of the integration of technologies such as IoT, 5G, AI, and big data, intelligent edge computing is rapidly transitioning from concept to reality. According to IDC, by 2025, about 75% of data will be generated and processed at the edge. The increasing volume of data generation also raises the demand for edge computing power. The intelligent edge will become a key computing area for end-to-end digital infrastructure in the digital age.

Combining Software and Hardware to Support the Ubiquity of Computing Power Networks

In the rapid development of intelligent edge computing, chips play a crucial role. “In fact, the innovative applications of intelligent IoT must be realized through underlying chips and supporting software. From the perspective of chips, as the demand for computing power increases, the requirements for chips will also become higher,” Zhang Yu pointed out.

Intel has long been deeply involved in the IoT industry, providing a diverse product line to meet user needs. At this year’s International Consumer Electronics Show (CES 2022), Intel released the 12th generation Core processors (codenamed Alder Lake S series and H series). This is Intel’s first processor series optimized for the edge, utilizing a high-performance hybrid architecture that integrates performance cores, efficiency cores, and a hardware thread scheduler (Thread Director) to optimize the acceleration of IoT application innovation, suitable for users in retail, manufacturing, healthcare, and video customer service for graphics, multimedia, display, and artificial intelligence computations.

Zhang Yu also emphasized that while Intel is committed to improving the performance of chip products, it also places great importance on achieving sustainable green computing concepts. “We cannot just emphasize the improvement of computing power; we must also recognize the importance of improving energy efficiency. Intel is always focused on promoting orderly social development in a sustainable manner.”

According to reports, from 2010 to 2020, the energy efficiency ratio of Intel’s Core product line improved by 14 times. From 2020 to 2030, Intel’s goal is to further improve the average energy efficiency ratio of its product line by 10 times. To achieve this, Intel has introduced the concept of “big and small cores” in the 12th generation Core processors, including performance cores and efficiency cores, utilizing different cores to match computing needs and achieve optimal loads to reduce overall power consumption.

Heterogeneous computing is also an important technical direction for addressing differentiated needs and improving processor energy efficiency ratios. Many loads have typical characteristics, whether in video conferencing scenarios, smart retail scenarios, or cloud gaming scenarios, the underlying technology involves video encoding, decoding, and transcoding. For these relatively fixed loads, using dedicated hardware for processing is more efficient and consumes less power. Therefore, in the Core product line, Intel integrated integrated graphics early on to form a hardware unit specifically for video encoding, decoding, and transcoding. Of course, for users with higher performance requirements, Intel can also provide solutions with discrete graphics cards. The overall approach is to improve energy efficiency ratios through heterogeneous architecture.

How to leverage these hardware products is also dependent on software support. Zhang Yu emphasized that a good software tool can enable developers to fully leverage hardware performance. Intel has been working with partners in this area, including the launch of oneAPI to achieve access to heterogeneous hardware resources through the underlying DPC++ open interface. For developers, using oneAPI allows for unified programming across different hardware.

Intel has also launched the OpenVINO toolkit to accelerate AI inference at the edge. OpenVINO enables developers to train in different AI frameworks. In some AI model scenarios, OpenVINO can provide up to 7 times inference acceleration, effectively enhancing work efficiency and increasing platform competitiveness.

Industry + Retail: Edge Intelligence Accelerates Implementation

With the development of intelligent edge computing, its successful applications have emerged in several fields.

Industrial manufacturing is one of the key areas for the penetration and development of intelligent edge computing. A recent research report from MarketsandMarkets shows that the global smart manufacturing market size was $88.7 billion in 2021 and is expected to reach $228.2 billion by 2027, with a compound annual growth rate of 18.5%. In this regard, Zhang Yu stated: “Machine vision, as an important technology for the intelligent transformation of manufacturing, is evolving rapidly with the rapid development of artificial intelligence, and its market size is growing rapidly. In fact, the current wave of artificial intelligence explosion is based on deep learning, which is based on image recognition. Currently, about 80% of the data collected in the IoT field is related to graphics and images.”

Based on this, Intel recently partnered with Xunbu Technology to launch a series of machine vision development kits. The development kits are framed by the Industrial Edge Insight Platform (EII) and visual software optimization package, integrating software tools and libraries such as OpenVINO, DPC++/C++ compiler, oneAPI Math Kernel Library (oneMKL), Vtune™ Profiler, IPP, OpenMP, and TBB, pre-setting typical industrial application scenarios and algorithm optimization reference methods to provide users with an end-to-end software framework for developing machine vision applications. Currently, the development kits have been successfully applied in industries such as automotive manufacturing, 3C/semiconductors, food packaging, and logistics warehousing, effectively aiding the manufacturing industry in its evolution towards intelligence and informatization.

The competition in the retail market is becoming increasingly intense, and the significant value demonstrated by artificial intelligence is gaining widespread attention from global retailers. Through the application of artificial intelligence technology, retailers can understand consumer preferences more effectively, providing personalized and unique services to enhance the attractiveness of retail services. At the same time, artificial intelligence also helps retailers automate more processes, achieving greater profitability. Data shows that by 2035, artificial intelligence will increase the profit margins of retail and wholesale businesses by nearly 60%.

In the face of this emerging application market, Intel, along with Hanshuo Technology and Microsoft, jointly launched intelligent edge solutions for the retail industry, primarily applied in smart shelf management and self-checkout loss prevention. The “people, goods, and venue” are the main elements of traditional retail, and this solution integrates Intel’s full-stack technology from software to hardware, helping retail customers build high-performance and easy-to-manage smart retail management systems.

In addition, edge intelligence has broad application potential in smart healthcare, smart transportation, and accelerating digital transformation for enterprises. “The development space for artificial intelligence in edge networks is immense, and market demand is very clear. Previously, people always thought that artificial intelligence was a high-end technology, with high thresholds and difficult deployment. However, the reality is that more and more companies have begun to deploy and operate artificial intelligence technologies successfully. This situation precisely illustrates that the barriers to using artificial intelligence are lowering, and it is also a typical example of the ubiquity of artificial intelligence,” Zhang Yu pointed out. In this process, Intel’s technologies, including the OpenVINO toolkit and a series of hardware products, also play a crucial role.

Localizing Adaptation to Break Through the “Last Mile”

In the process of accelerating the implementation of intelligent edge computing, how to better adapt it to local needs and break through the bottleneck of the “last mile” is also a very critical aspect. In an interview, Zhang Yu specifically pointed out: “The way technology is implemented in data centers and at the edge is different.” After all, platforms, computing power, and conditions are different, and the implementation methods will not be the same. This means that when artificial intelligence is implemented at the edge, specific issues that only arise at the edge must be resolved.

For example, artificial intelligence generally consists of two major stages: the training stage and the inference stage. The training stage requires a large amount of data for training, and prior to training, data must be annotated, marking the objects that need attention in the data, and then the annotated images are transmitted to the training platform for training, ultimately generating an AI network model that can perform inference.

However, this model may encounter challenges in edge applications. In reality, there often isn’t enough data available for training at the edge. For example, defect detection is one of the most common applications of artificial intelligence in smart factories. However, the probability of defects occurring on a normal production line is not very high. This means that the sample data available for training is quite limited. How to train a usable model with a small sample is a challenge that intelligent edge computing needs to address.

Furthermore, during training, annotation is required, but in edge applications, the personnel operating artificial intelligence are often the production staff on the production line. These individuals often find it difficult to have extra energy to perform this annotation work. Therefore, developing automated annotation tools to reduce the burden on operators and ensure that the closed loop of training, annotation, and inference can truly operate is another challenge that Intel needs to solve when implementing artificial intelligence at the edge.

“In fact, many issues at the edge need to be solved in a specific manner,” Zhang Yu emphasized.

Building an Industrial Ecosystem to Address Fragmentation Challenges

Of course, in the field of intelligent edge, all of Intel’s work relies on the support of partners. “Intel has long been deeply involved in the IoT industry, providing a rich variety of product types, and we hope that through these different combinations of products, our partners can have greater room for development in the IoT field,” Zhang Yu pointed out.

When talking about Intel’s competitive advantages in the field of intelligent edge, Zhang Yu attributed it to having a broad partner base. “This allows us to understand user demands more profoundly, and based on these demands, formulate software and hardware solutions. Whether these users are at the network front end, edge, or data center, Intel can provide end-to-end solutions that align with the trends of digital infrastructure development.”

Although the IoT has limitless possibilities, the fragmentation problem remains a significant challenge for vendors. In this regard, Zhang Yu stated that on one hand, we need to ensure that our products are as universal as possible; but on the other hand, project implementation always relies on close cooperation across the entire industry chain.

Different markets have different needs. The characteristics of the Chinese IoT market are rapid development and many innovative customers. These customers may not be large in scale, but they possess strong innovation capabilities. In recent years, Intel has worked extensively with domestic partners, including system integrators, to integrate the product technologies from different segments of the industry chain to solve fragmentation issues.

In 2016, Intel and its partners jointly established the Edge Computing Industry Alliance. This is currently the largest alliance organization in China focused on edge computing, with over 300 member units, covering chip manufacturers, system integrators, software developers, ODMs, OEMs, and more. Through this alliance platform, we collectively explore solutions for IoT and edge computing and industry solutions to tackle fragmentation issues.

“Water benefits all things without contention” may be the best description of Intel’s positioning in the entire ecosystem. For many years, through silent work, Intel has connected all segments of the edge intelligence industry chain, including original design manufacturers (ODMs), original equipment manufacturers (OEMs), system integrators (SIs), and independent software developers (ISVs), providing the industry with an end-to-end solution oriented towards digital infrastructure.

Intel CEO Pat Gelsinger proposed the concept of four super technological forces at the “2022 Intel On Industry Innovation Summit” held this year, including ubiquitous computing, ubiquitous connectivity, artificial intelligence, and infrastructure from cloud to edge. Zhang Yu emphasized that these four super technological forces are, in fact, the four pillars required to build a green, intelligent, end-to-end digital infrastructure in the future. If these four super technological forces are used comprehensively, what is built is an end-to-end digital infrastructure. In the future, this digital infrastructure will support more new creations and applications. These infrastructures have two characteristics: one is intelligence, and the other is sustainability.

For submissions, please click “Read Original”