As usual, before I begin, I do encourage my readers to carefully do their own research. I am not an electrical engineer, computer engineer, or physicist.
Apple Silicon, the coined term for their new lineup of extremely powerful Chips (the M1, the M1x, the M2), is not entirely revolutionary. Apple Silicon is Apple’s implementation of an ARM (Advanced RISC (Reduced Instruction Set Computing) Machines).
ARM Chips have a number of advantages. Although they were built to reduce power consumption, the outcome has had a manifold effect. Because of reduced circuitry (one of the key components of ARM) & the localization of components (creating shorter distances for electrical charge travel), ARM-based chips are incredibly fast and efficient.
Currently, 86x based architectures of CPUs dominate the market. This has presented a slight problem for ARM based chips because the software that is assembled for 86x architecture is not compatible with ARM based architecture. To illustrate an example of how these architectures differ, let us imagine that instead of computer architectures, we looked at “life architectures.”
For the following two diagrams:
BLUE represents tasks to be done
GREEN represents how we can go about those tasks
ARROWS represent distance & time to accomplish tasks
Visually the 86x Architecture relies on parts of the computer that are semi-randomly sourced and implemented per manufacture (from a market-sector perspective). For example, most 86x based chips are made by Intel and bought from computer manufactures like Apple, Microsoft, & Dell. The manufacture must source other essentials such as GPUs, Memory, Storage, Batteries, etcetera, and connect all of these things together on the “motherboard” which is a circuit-based housing specifically outfitted for integrating computer parts — usually individualized by each manufacture.
Right now I am writing this on an Intel-based computer, and to be honest, the performance is not bad, it’s actually pretty good. The diagram above (x86) is fundamentally less organized and less efficient, but it does work well… just not as well as ARM has the potential for.
An ARM-based architecture might look something like the graphic directly above. The core components being localized reduces distance and complexity for any given connection. This allows quicker communication and dispatching while reducing energy consumption. From a motherboard design perspective, there is less work to be done to cool the unit as ARM-based architectures are localized and create a smaller target zone to cool. The main disadvantage of ARM-based chips is two-parted: (a) because the circuitry is fundamentally different, 86x assembled software is incompatible, and (b) they are not customizable or upgradable.
Conclusion: Apple Silicon uses ARM based architecture for a number of reasons.
A) In house manufacturing of the components essential to their computer line reduces costsB) Having all of the essential components built within one chip further reduces costsC) Being able to design the entire computer around in house hardware reduces costs, by reducing design and manufacturing complexityD) ARM based performance and efficiency enchancements help them maintain and improve upon four categories: 1) Longer Battery Life 2) Increased Performance 3) Higher Profit Margin 4) Lower end-consumer cost
It’s not “witchcraft” because it utilizes simple physics and organizational methods that have been around for ever.
Proof: I wish to narrow in on how the physics of this works and how it is not entirely new. I am a fan of reflexivity within things, enough so I could go all meta, but for this article I just want to stick to water and electricty — and how they are inherinently similar — to describe why shorter distances, reduced circuity, and reduced resistance found within ARM based architectures lead to the performance and efficiency improvements that they do.
How does water pressure work?
To understand water pressure, more specifically water pressure in a closed loop system, I would like to explain a couple of components.
First, how does pressure work? Well pressure is equal to the force applied over the area in which that force is applied. A bullet and car applied to a gelatin figure with the same amount of force will cause different outcomes (Note: force & velocity are different). Both the bullet and the car will undoubtly create equal and opposite reactions upon the gelatin figure and themselves, but only the bullet will rip through the figure because the implied impact surface area of the bullet is much smaller than that of the car. Fluid pressure is not inherently different. Key takeway: surface area contributes to pressure.
Second, fluid pressure (atmospheric, gas, liquid, eccetera) distributes within the volume of space that they are being contained in. There are a number of forces you could account for such as surface tension, torsion, bouyancy while taking into account density, mass, surface friction, eccetera. However, if you have ever been swimming, you will note that there is a larger amount of pressure applied to your body when submerged that is not present when not submerged. This is because the atomospheric fluid is less dense than water, but for this example lets imagine water is the only fluid in play. Water, like other fluids, distributes evenly accross all surfaces. Before you enter the pool, the water is pushing against the barriers that contain it evenly. As soon as you enter, water rises because of an increased volume and distributes evenly across all surfaces as before, except now you are a surface to distribute on. This is illustrated by figure (b) below (ignoring the force diagram for net bouyant force). Key takeaway: pressure distributes evenly.
In a closed loop system, like the figure below, the pressure distributes outward similarly to figure (a) above with the arrows pointed towards the outer ring. The pump in the diagram below creates an “artificial” pressurized system. The surface area of the interior of the pipes helps handle this pressure. For example, a pipe of 4-inch diameter would require a stronger pump because of the two previous examples. While a pipe of 0.5-inch diameter would require a less powerful pump. Similarly if the pipe stretched a kilometer versus a meter, the pump would have to be stronger in the former case to match the pressure of the latter case. Key takeway: amount surface area and amount of “artifial” force contribute to overall pressure.
Note: “artifial” is used to define man-made contributions
So, cool… fluid pressure, surface pressure, and closed loop home water systems. What does this have to do with anything about ARM chips? Well, electricty in a closed loop circuit distributes the charge evenly across the circuit (ideally). The charges travel to the “path of least resistance.” Wiring for electric circuits contain n amount of “resistance.” A circuit without a flow of charges, is similar to a closed loop home water system without the flow of water. When electric or water pressure enters a closed loop, the charges or fluid try to distribute evenly finding the “paths of least resistance”. The components that make the distrubition more efficient have to do with the amount of distance needed to travel and the amount of resistance the charge or fluid faces.
ARM based chip architecures decrease distance and in some cases decrease resistance needed because of their localization.
It’s not “witchcraft” because everything I have describe above is not new knowledge. The first civilized application of the control of water was accomplished by the Roman Civilization in 312 B.C. via an aqueduct. The first use of electricty was by Old Pops Benjy. And the study of these things, our reality — how to utilize it and improve it — have been around since mankind itself. Apple is simply being smart.
Enjoy the read?
Leave a clap or comment
Share with friends or on your favourite social platform
