How a CPU Turns 1s and 0s Into Everything You See
video description
Date: 2026-10-28
Related videos
Comments and reviews: 20
randyscorner9434
Very nice. Covered a lot of ground in just a few minutes. I designed processors at Intel starting in 1980 through mid-2000s. One of my designs was the FPU in the 387, 960KB, and 486 products. We had a 68-bit adder for the mantissa of the floating point unit and the carry chain on it was one of the primary critical paths of the chip, which determined the maximum frequency we could reach at the time. You got all the way through cache hierarchies, which took an enormous amount of effort and genius to get right given the many cache coherency issues. And we did the first glueless multiprocessor, which allowed a second (and more) processors to be attached directly behind an L3 cache. The microarchitecture of processor stalled out after the 90's and 2000's which meant that the number of cores increased to add performance. One of our innovations was hyperthreading, or running two separate threads on the same hardware, giving about a 1. 7X performance boost with a small number of additional transistors. It's out of fashion now, but so is more microarchitecture for faster single cores. It was and is a great ride to work on these things.
reply
Very nice. Covered a lot of ground in just a few minutes. I designed processors at Intel starting in 1980 through mid-2000s. One of my designs was the FPU in the 387, 960KB, and 486 products. We had a 68-bit adder for the mantissa of the floating point unit and the carry chain on it was one of the primary critical paths of the chip, which determined the maximum frequency we could reach at the time. You got all the way through cache hierarchies, which took an enormous amount of effort and genius to get right given the many cache coherency issues. And we did the first glueless multiprocessor, which allowed a second (and more) processors to be attached directly behind an L3 cache. The microarchitecture of processor stalled out after the 90's and 2000's which meant that the number of cores increased to add performance. One of our innovations was hyperthreading, or running two separate threads on the same hardware, giving about a 1. 7X performance boost with a small number of additional transistors. It's out of fashion now, but so is more microarchitecture for faster single cores. It was and is a great ride to work on these things.
reply
joegagliardi3984
TSMC in Taiwan is sitting pretty. They now control between 70% to 72. 5% of the global semiconductor market. It’s their business model that has made them so successful. They’re a pure-play contract manufacturer, meaning they ONLY manufacture chips for other companies (like Apple, Nvidia, and Qualcomm) rather than designing its own branded products. This strategy reassures clients that TSMC would never steal their intellectual property or compete with them.
On a more pessimistic note, it’s possible that the semiconductor monopoly Taiwan enjoys will lead to the start of WWIII, when China finally decides to take the island by force.
a chip Sam’s Clubthat the world’s major companies need or major clients like Nvidia, Apple, AMD, and Qualcomm
reply
TSMC in Taiwan is sitting pretty. They now control between 70% to 72. 5% of the global semiconductor market. It’s their business model that has made them so successful. They’re a pure-play contract manufacturer, meaning they ONLY manufacture chips for other companies (like Apple, Nvidia, and Qualcomm) rather than designing its own branded products. This strategy reassures clients that TSMC would never steal their intellectual property or compete with them.
On a more pessimistic note, it’s possible that the semiconductor monopoly Taiwan enjoys will lead to the start of WWIII, when China finally decides to take the island by force.
a chip Sam’s Clubthat the world’s major companies need or major clients like Nvidia, Apple, AMD, and Qualcomm
reply
disjustice
Little bit of a quibble with how you presented GPUs - displaying simple pixel raster data as you presented in your example does not require billions of operations for each frame. It's calculating that raster image based on the geometry and lighting data in a 3d model of a scene that takes lots of small parallel calculations. Really great intro to computer/CPU architecture though!
reply
Little bit of a quibble with how you presented GPUs - displaying simple pixel raster data as you presented in your example does not require billions of operations for each frame. It's calculating that raster image based on the geometry and lighting data in a 3d model of a scene that takes lots of small parallel calculations. Really great intro to computer/CPU architecture though!
reply
RipRoarLB
I've seen so many of these and in high level context, everything make sense, but it really does not tell you at the core, how it works. How does a computer now a t is a t, and how does the computer and the core level figure out a complex If statement in Excel, or multiply huge numbers, sort, at the core level of the computer. You say the computer organizes the memory. ., but how!
reply
I've seen so many of these and in high level context, everything make sense, but it really does not tell you at the core, how it works. How does a computer now a t is a t, and how does the computer and the core level figure out a complex If statement in Excel, or multiply huge numbers, sort, at the core level of the computer. You say the computer organizes the memory. ., but how!
reply
stabilini
In these highly advanced times, there is something very endearing about seeing how logic gates work. It is a bit mind-boggling to think that every CPU is operating millions of times per second, even when you aren't doing anything with the computer. Side note: an actual nVidia GPU drawing chip (in GDSII format file) could be as large as 100 GB.
reply
In these highly advanced times, there is something very endearing about seeing how logic gates work. It is a bit mind-boggling to think that every CPU is operating millions of times per second, even when you aren't doing anything with the computer. Side note: an actual nVidia GPU drawing chip (in GDSII format file) could be as large as 100 GB.
reply
okithdesilva129
One of the great things about these videos is that Paul starts from the absolute basics like Current is the flow of electrons. It sometimes makes you remember something you have probably forgotten and also clear out doubts. That's a reason why these videos are so popular among the STEM community.
reply
One of the great things about these videos is that Paul starts from the absolute basics like Current is the flow of electrons. It sometimes makes you remember something you have probably forgotten and also clear out doubts. That's a reason why these videos are so popular among the STEM community.
reply
mrkkitson
Very good! I learned about logic gates in the Marines when I was trained on an old system that did tactical air traffic control. Learned more about computing systems with IBM as a customer engineer. I became a mainframe specialist. Your explanation is exceptional, probably the best I've ever heard
reply
Very good! I learned about logic gates in the Marines when I was trained on an old system that did tactical air traffic control. Learned more about computing systems with IBM as a customer engineer. I became a mainframe specialist. Your explanation is exceptional, probably the best I've ever heard
reply
Latrocinium086
Hate to be the nay sayer, but there’s a big logical gap from the adder to the registers and instruction sets.
How does an adder make a register or how is that logic executed What is the circuitry that makes those decisions
Could probably be a video all on it’s own.
reply
Hate to be the nay sayer, but there’s a big logical gap from the adder to the registers and instruction sets.
How does an adder make a register or how is that logic executed What is the circuitry that makes those decisions
Could probably be a video all on it’s own.
reply
USMC1984
61yo here and have never had a computer course, I’ve read and watched many tutorials online about computers but never really understood how they operated, this one video has cleared many questions. Awesome video!
I’m going to share this with many others.
reply
61yo here and have never had a computer course, I’ve read and watched many tutorials online about computers but never really understood how they operated, this one video has cleared many questions. Awesome video!
I’m going to share this with many others.
reply
jaglover
This brought me back to my Computer Logic and Organization 201 class in college 40 years ago. The only difference is we didn't have SSD or all the caches.
It always amazed me that everything a computer did boiled down to load register and store register.
reply
This brought me back to my Computer Logic and Organization 201 class in college 40 years ago. The only difference is we didn't have SSD or all the caches.
It always amazed me that everything a computer did boiled down to load register and store register.
reply
BlackbodyEconomics
Two beautifully bagay books are CODE (by Charles Petzold) and Digital Logic Design by. somebody.
(can't remember what bagay is in English. sorry. Bagay means they go well together. They compliment each other. They're bagay)
reply
Two beautifully bagay books are CODE (by Charles Petzold) and Digital Logic Design by. somebody.
(can't remember what bagay is in English. sorry. Bagay means they go well together. They compliment each other. They're bagay)
reply
slayzye
You're an absolute legend. They teach us these things in college as subjects and we rarely are given a proper explanation as to how stuff is connected but you in just 15 minutes made it so easy to understand. Thank youuuu
reply
You're an absolute legend. They teach us these things in college as subjects and we rarely are given a proper explanation as to how stuff is connected but you in just 15 minutes made it so easy to understand. Thank youuuu
reply
okithdesilva129
A request. Can you consider continuing the series on electric motors, electronics and electrical engineering with videos on things like capacitors(like the one on resistors) and about synchronous motors Thank you!
reply
A request. Can you consider continuing the series on electric motors, electronics and electrical engineering with videos on things like capacitors(like the one on resistors) and about synchronous motors Thank you!
reply
Henreys
This explains well how simple on/off switches build logic gates, then adders, then whole cores. The cache hierarchy walkthrough is the best part. GPUs matter for AI because their parallel design fits repetitive math.
reply
This explains well how simple on/off switches build logic gates, then adders, then whole cores. The cache hierarchy walkthrough is the best part. GPUs matter for AI because their parallel design fits repetitive math.
reply
engineering_mindset
Fantastic explanation. I doubt I'll ever use this day to day in my ammonia refrigeration job, but it's nice to know what's going on in the background when I'm working on programs interacting with the PLC.
reply
Fantastic explanation. I doubt I'll ever use this day to day in my ammonia refrigeration job, but it's nice to know what's going on in the background when I'm working on programs interacting with the PLC.
reply
InsideChipFactory
I’ve always wondered how a bunch of 1s and 0s can create everything we see on a screen. It’s amazing how the CPU turns such simple signals into something so powerful!
reply
I’ve always wondered how a bunch of 1s and 0s can create everything we see on a screen. It’s amazing how the CPU turns such simple signals into something so powerful!
reply
monshome373
Writing assembly code makes you more familiar with inner works of CPU, but your brain goes through much pain to reach the point when the code is flawless.
reply
Writing assembly code makes you more familiar with inner works of CPU, but your brain goes through much pain to reach the point when the code is flawless.
reply
OydaOydaasdsa
Thank you. I like this video. Please continue this (CPU) series. I would pleased to see CPU pipelines, OoO execution and similar essential topics.
reply
Thank you. I like this video. Please continue this (CPU) series. I would pleased to see CPU pipelines, OoO execution and similar essential topics.
reply
UKDeeGee
Glazed over after 5 mins. Yep, none the wiser. Computer geeks just can’t seem to find a way to explain it. The question goes unanswered, again
reply
Glazed over after 5 mins. Yep, none the wiser. Computer geeks just can’t seem to find a way to explain it. The question goes unanswered, again
reply
ElectroAtletico
IBM recently announced a prototype 0. 7mm (density node) chip that packed nearly 100 BILLION TRANSISTORS onto a fingernail-sized chips.
reply
IBM recently announced a prototype 0. 7mm (density node) chip that packed nearly 100 BILLION TRANSISTORS onto a fingernail-sized chips.
reply
Add a review, comment
Other channel videos
