At WWDC 2020, Apple announced the direction of integration between Apple Silicon and all Apple OS. In other words, Apple has predicted a new challenge with high risk. And the keynote is expected to ignite the long-running debate between the CISC and RISC camps once again.
The development of computer technology was a remarkable development in my daily life. The living and breathing dinosaurs in the movie Jurassic Park were realistic in the picture of the encyclopedia. The experience of walking and running together as Super Mario in the flat world evolved into Super Mario in 3D interaction was amazing. Curiosity about the principle was directed to a company called Silicon Graphics. Armed with the branding of a reality series using UNIX, it was a terrifyingly priced machine that led both content and devices simultaneously to innovate. (Machines at the time are still traded by eBay’s Harbists and users are shared through YouTube.) It is surprising to see now that the same chips are arranged in parallel like a building. Then, there were early 3D consoles that could directly interact with immersive experiences such as movie experiences at the time, and unknowingly, RISC CPU was put in a corner of the promotion for children’s targets. The question of what RISC is was difficult to experience and judge in the CISC-based Windows+Intel world.
A new wave of Phono-sapiens introduced the word ARM to the public, and the word ARM again aroused curiosity about RISC. In the daily life of CISC, the parallel of RISC emphasized the meaning of the usability of coexistence rather than superiority.
I would like to talk about the two theoretical differences that are difficult to talk about specifically as the difference between the horizontal and vertical processing of the problem array. The CPU was the object of thinking about using a computer-understandable method to plan and process the necessary commands on its own. CISC has maintained its philosophy of creating new instructions and processing them vertically in length as a benefit of variable processing when one CPU becomes more experienced. RISC should design complex command-splitting and abstraction rules (similar to farming, planting in parallel to traditional instrument sounds). The results of execution (although difficult and complex planning must precede), are cost-effective for RISC. And the era has come when most split abstract commands are solved by runtime. Furthermore, through various research and development in the non-memory field, various problem-solving processing units have come to the market. With the help of the surroundings, vertical problem solving was turning into an era that could be solved horizontally. The electricity and its path for processing commands as a bonus are drawn on silicon, which also had a physical speed limit of electricity and silicon. Unless the era in which quantum computing is possible for the general public has come, parallel processing competition in processing speed competition has been inevitable.
If parallel processing is performed efficiently, the world can be created more densely and efficiently. In comparison, it seems similar to introducing the concept of blocks in cities and achieving more efficient lives. On the contrary, vertical allows for less error-prone accuracy and efficient response to repetitive and regularized problems, and has the advantage of relatively shorter execution time than planning time.
Back in the keynote, I felt that the computer, a command-processing machine for solving problems, repeatedly competing and challenging derivatives and parallels in horizontal and vertical ways was valuable as a series of processes that developed into effort and innovation in problem solving.
Whether RISC is actually an excellent design, Apple’s decision is expected to confirm this. But it’s a little more complicated. Apple’s chip may actually prove to be excellent, but it may not have much to do with ARM-based RISC. So it became a keynote part that is more anticipated for release.