Learn computer architecture online: four hands-on options
The best way to learn computer architecture is to build a computer, and four well-known options let you do that online: the Nand to Tetris course, MIT's 6.004 Computation Structures on OpenCourseWare, Ben Eater's 8-bit breadboard computer videos, and digiwleea. They differ in how deep they go, how you build, and how much theory comes with it. Here is what each one covers, checked on its own site.
What these courses have in common
All four start from logic gates (or below) and work up to a processor, so the CPU stops being a black box. The more you build yourself, the better it sticks: every option here has you make the parts, not just read about them. (digiwleea is an independent project, not affiliated with any of the courses below.)
Where digiwleea fits
digiwleea is the most hands-on and the most beginner-friendly of the four: it runs in the browser with nothing to install, starts from CMOS transistors, and checks every part you build against its truth table, through a 40+ level course to an 8-bit CPU you program in assembly, with free theory lessons and an AI tutor (Wleea). It does not go as far as MIT 6.004, which also covers pipelining and virtual memory, or as far up the software stack as Nand to Tetris, which continues to a compiler and an operating system. Many learners use digiwleea first and go on to those.
The options, one by one
- Nand to Tetris (Coursera course by Shimon Schocken and Noam Nisan, Hebrew University of Jerusalem; book: The Elements of Computing Systems, MIT Press): projects build logic gates, an ALU, RAM, a CPU and an assembler on a hardware simulator, ending with a general-purpose computer called Hack. The lectures, project materials and tools are on nand2tetris.org. Compared with digiwleea.
- MIT 6.004 Computation Structures (MIT OpenCourseWare, Spring 2017, Chris Terman; openly licensed): an undergraduate course on the architecture of digital systems, from gates and instructions up to pipelined systems, instruction set design and virtual memory, with lecture videos and notes.
- Ben Eater's 8-bit computer (videos at eater.net/8bit): build an 8-bit computer on breadboards from 7400-series chips: clock, registers, ALU, RAM, program counter and control logic. Compared with digiwleea.
- digiwleea (in the browser; the lab and Act I are free, Premium for the full course): a 40+ level course from CMOS transistors to an 8-bit CPU, with theory lessons and an AI tutor. How to build a CPU from scratch.
Which one to pick
For a university-level course with the most theory, MIT 6.004. For the full path from gates to software, Nand to Tetris. For real hardware on your desk, Ben Eater's breadboard build. For the gentlest start, in the browser, from the transistor up, digiwleea.
No sign-up is needed to start: open the lab and place your first transistor, or read the free theory course. The account (and the AI tutor) are optional.
See for yourself: build a logic gate from transistors in the simulator, then follow the path from sand to an 8-bit CPU.
Open the digiwleea lab →Frequently asked
What is the best way to learn computer architecture?
By building a computer: start from logic gates, combine them into adders, registers and memory, and finish with a processor you can program. Nand to Tetris, MIT 6.004, Ben Eater's 8-bit computer and digiwleea all follow that path at different depths.
Is there a free computer architecture course?
Yes. MIT 6.004 Computation Structures is on MIT OpenCourseWare under an open license, the Nand to Tetris materials are on nand2tetris.org, Ben Eater's videos are online, and digiwleea's lab and first act are free in the browser.
Do I need to know electronics first?
No. All four start from logic, and digiwleea starts from the transistor as a simple switch, so no electronics background is needed.
How long does it take to learn how a CPU works?
A focused course takes weeks, not years: the Nand to Tetris hardware part is listed on Coursera at about 10 hours a week for four weeks, and digiwleea's 40+ levels from transistor to CPU can be worked through at your own pace.