In this blog post, we’ll explore the concept and structure of the application processor (AP), which handles various smartphone applications.
These days, cell phones have become an indispensable necessity for modern people. In South Korea, in particular, the number of smartphone users has grown significantly, surpassing 15 million, and most people around us now own a smartphone. Smartphones use touchscreens as their primary interface to run a wide variety of applications; they enable not only information searches via wireless communication but also a variety of simple games using touch controls. In addition to simple games, there are even racing games where you rotate the smartphone as if you were turning a car’s steering wheel. What exactly enables smartphones to run these countless applications?
Many people would likely think of a computer’s CPU. Computers, like smartphones, can run a wide variety of programs with ease, and the main component that makes this possible is the CPU. However, there is a significant difference between a smartphone’s processing unit and a computer’s CPU. A CPU is a general-purpose processor, which prioritizes versatility and focuses on handling a wide range of functions. Consequently, while it can perform various tasks, its performance per task is relatively lower. In contrast, smartphones use an Application Processor (AP) to perform various functions, and the AP is optimized for specific applications tailored to its purpose. For example, a video engine processes video data, while a graphics engine processes graphics data. Consequently, although the range of functions it can perform is narrower than that of a general-purpose processor, it can handle specific functions more efficiently. Since smartphones perform a more limited range of functions compared to computers, they rely on APs. How are these APs constructed? Before exploring this, let’s first examine the “chip”—the fundamental concept behind the AP—and the “transistor,” the chip’s basic component. A transistor is a fundamental device in electronics; depending on the voltage applied to its input, the voltage at its output represents either a 0 or a 1. This allows us to generate codes such as 010100011 or 0000101011, which are fundamental to computers. A chip is an arrangement of countless transistors designed to produce the desired results; however, because an enormous number of transistors are required, it is impossible to manufacture them by hand. Therefore, hardware programming is performed using a computer; hardware programming is the process by which, just as with software development, specific code is entered, and the computer automatically arranges the transistors accordingly. An AP consists of multiple chips, and there are two methods for configuring an AP. System-on-Board is a method of connecting multiple chips on a single board to form a single system. For example, if System Company A wants to build a certain system, it involves appropriately connecting chips from Company B, Company C, and Company D to achieve the desired system. Conversely, System-on-Chip is a method of implementing all system functions within a single chip. For example, if System Company A wants to build a certain system, it constructs the system using a single chip through direct hardware programming. The advantage of the former is that design and modification are relatively easy, whereas with the latter, once a chip is manufactured, it cannot be modified and must be redesigned. The advantage of the latter is that the system’s price is lower; although the price of a single chip is somewhat higher, fewer chips are required. Recently, the System-on-Chip approach has been the predominant method.
