|Jin-Hyo, Park Bio
SK Telecom CTO, Head of ICT R&D Center
5G for Hyper-connected Society
|In the era of the Fourth Industrial Revolution, everything is connected and analyzed. It is also believed that the cyber space may not be discerned from the real world eventually. However, explosive data growth expected by the Fourth Industrial Revolution cannot be handled by existing network infrastructures. This is where the 5G mobile communication network plays and why 5G is called a “steam engine of this age.”
5G technology is characterized as ultra-fast speed, ultra-low latency, and hyper-connectivity. A physician on the land can operate on a patient at a medical facility in an island village through a hologram delivered on 5G networks. And 5G technology realizes offline services in the virtual world such as finance, shopping, and performance that has been provided in the real world. In addition, ultra-high-quality media such as UHD and VR / AR are delivered to users through 5G networks as if they are actually seen.
It is also expected to evolve into a hyper-connected society developed through 5G and IoT technologies. For constructing the hyper-connected society, the smart city and the smart factory are essential components. The 5G infrastructure built in the smart city delivers traffic signal and intersection conditions to autonomous vehicles, letting them know which lane they should move to. Additionally, IoT devices detect accidents and disasters in the city and simultaneously spread the disaster situation through the 5G network, greatly reducing the risk. And the smart factory can replace human labor with machine operation through unmanned process in factory for citizen to concentrate on creative work.
In a world where everything is connected, information security threats must be carefully addressed. Attacks on traffic, electricity, and communication infrastructure control systems in the smart city will pose a serious threat to social safety. And if manufacturing information and systems in the smart factory are attacked, corporate losses will be astronomical. Therefore, strong communication security by 5G infrastructure will be a basis of the bright future of the hyper-connected society.
|Seok Hee Lee Bio
SK hynix Inc., President & Chief Operating Officer
Memory-Driven Computing: The Future Evolution of Cognitive IoT Technology
|As the ICT industry has entered the era of the Fourth Industrial Revolution, the “second wave” of artificial intelligence (AI) is also sweeping across the industry.
In the second wave of artificial intelligence, which is characterized by more sophisticated AI services, there are growing needs for faster and more accurate and effective data processing technologies. Some of those technologies that recently gained traction are Cognitive IoT technologies that takes advantage of Cloud with robust learning algorithms such as Machine Learning and Deep Learning. It also adopts intelligence in IoT with highly sophisticated sensors and brain-like processors for implementation mainly in Autonomous Vehicle, Smart Factory, and Smart Home. In this trend, memory devices are expected to play a key role as AI environment becomes more ‘Memory-Driven Computing’. Therefore, in this presentation, I would like to shed light on the past and present of Memory Device technology under the AI and ICT environment that is becoming increasingly sophisticated, and also talk about the future memory device technology.
In traditional and current systems, memory devices have been considered a fundamental performance bottleneck. As we approach the end of traditional Moore’s Law, however, performance gain for CPUs from technology scaling is slowing down, raising questions on the scalability of the performance of a computing system. In future emerging systems, memory devices will be a key factor for improving performance. DRAM has now started contributing to performance improvement in terms of high bandwidth through HBM. On the other hand, the NAND Flash industry has successfully commercialized 3D NAND. Besides, emerging memory technologies such as PCRAM, STT-MRAM, and ReRAM are being prepared for adoption in the near future. These memory devices will provide significant value to different applications based on their system-level workload, especially in the era of data-driven artificial intelligence computing environment, leveraging their own differentiated advantages in terms of performance, capacity, reliability, and cost.
In the second wave of AI, Cognitive IoT is an important link that enables all objects to communicate with each other and create new value, and the basis of this will be “Memory-Driven Computing”. Also, changes in an AI environment will be an inflection point for memory device technology in terms of business model, technology innovation strategy, and ecosystem collaboration, and such changes will create challenges as well as opportunities for us. At the same time, in the second wave of AI, memory device is expected to play a key role as it positions itself higher up in the value chain through ‘forward integration’ which adds value through value chain expansion (Device to System) and customized standardization.