Same but I’m a CE student

For real... now I understand M1 much much better.. it was mystery to previously

I agree. Best m1 video yet

AGREE!!

Computer science name is a open certificate and much people in internet doing internet works after having other certificate so anyone can inter this .

Thanks Paul for the kind words! I’ll do my best to keep bringing you great content! Thanks for taking the time to watch.

really good!!!

@Igor Velky You are out of context and sound butt hurt that Intel is STILL behind since Broadwell. Deal with it.

As a layman, I truly agree with this assertion! Keep up the good work mate!

Great job! CISC is doom!

Steve jobs never designed a chip and M1 is not simple. How do you move mountains with simplicity?

@@Teluric2 my comment is about the video explanation, not about designing or making the chip

if thats the case do CISC vs RISC also then

Chip instructions are sort of like pixels on a computer image- they are the most basic units that programs can be broken up into. They are also hard-coded into the physical structure of the chip. Each instruction is its own series of transistors and circuits that can complete that particular instruction. ARM and x86 are chip instruction architectures that have their own sets of instructions, and their own approaches to making instructions.

ARM and x86 belong to two different classes of instruction architectures. ARM is a RISC architecture and x86 is a CISC architecture. They are the most widely used of their kinds, but not the only examples of each. Neither are static- they develop a larger number of instructions over time. RISC architectures use instructions that are all of the same length, while CISC architectures can use instructions of a variety of lengths. Doesn't seem like such a big difference, but it means that ARM (a RISC architecture) can more efficiently manage computer resources. It took a lot longer to mature because it's fixed instruction length meant that it was difficult to implement enough instructions to do complex tasks- and it uses more memory resources. In the modern era though, those issues aren't such a big deal. It's complex enough now to do anything you'd want it to do, and memory is a lot cheaper today. The power and heat efficiency also means that you can run those chips faster. It's advantages have been especially useful in small electronics- an increasingly expanding use case- and its development has seen it used in more complex applications.

You can kind of think of the difference between the two types of architectures as a game of tetris. CISC is a regular game of tetris- all the shapes that fall down are irregular in shape and unpredictable- but with a bit of strategy, you can fit it all together and eliminate a row. The faster it goes, the harder it is to strategize quickly enough. Getting a high score is hard. RISC architectures are like playing a modified version of tetris where all the shapes are the same 2x2 block one after another. It would be a lot easier to get a high score, huh? There'd almost be no difference between a novice and a master. And the game would have to go A LOT faster before you got overwhelmed and failed.

Or at least, that's my understanding of them. I could easily have gotten something wrong.

Totally agree. It's the best explanation I've seen and one that needs to be seen by others. His critical points demonstrate how out of date AMD and Intel chip designs are compared to the M-series and why Apple's corporate decision to vertically integrate everything is the only way to achieve ultimate speed. Yes, there will be custom PCs that could be made faster than Macs but nothing off the shelf will be comparable. Apple's design is so much better and faster that they will be able to blow anything away from any other mainstream computer vendor. The first three M-series devices are entry level, yet they have the fastest single CPU Geekbench scores (not counting fake Ryzen-based iMacs which are being operated outside the legal use documents) with multi-core and GPU results being more than decent when considering how inexpensive they are. I'm sure we'll all be blown away when the next version of the M-series is released. Single core might not be much faster (could be with increased temperatures) but multicore and GPU results could more than double if not quadruple the initial results. I've seen people say a 32 core M-series could cost $30K but I disagree. The M-1 SoC probably only costs Apple at most $200 so a 32-cores might only cost them less than $1000. I may be way off but if I'm not we could see an iMac and Mac Pro really dropping in cost without hurting Apple's profit margin.

it seems like he got all the details from a blog post i read a few weeks back which is eerily very similar to how he is explaining it even the step by step process

@@delta0307 do you have a link to it. Must admit the video format is sometimes more convenient. BTW he does link to a written article on the same topic in the description.

@@AndreJohnMas Try typing the title in google. There should be an artible/blog from dubugger.medium

For anyone else: debugger.medium.com/why-is-apples-m1-chip-so-fast-3262b158cba2 (assuming this was it).

Watch this space