Build your CS foundations
Follow a program from execution to memory, complexity, and the CPU. Use computer science foundations to answer questions that come up while building software.
COMPUTER SCIENCE, CONNECTED
Go beyond code that works.
Follow a program through memory, the CPU and the OS to understand what happens inside a computer.
Use lessons, practice questions and demos without signing up.
let total = 0;
for (const value of values) {
total += value;
}Look at one computation through several layers.
Study 13 foundation textbooks and 528 other materials in Japanese or English. Check foundations → · Predict and experiment →
CHOOSE YOUR STARTING POINT
Try a short session with a lesson, practice questions and a look at the mechanism.
Follow a program from execution to memory, complexity, and the CPU. Use computer science foundations to answer questions that come up while building software.
Study processes, virtual memory, and concurrency in sequence. Build the foundation needed to understand network communication and databases.
Start with complexity, memory hierarchies, CPUs, and concurrency before studying AI execution systems. See where model theory fits into the wider learning path.
Inspect internal state and connect the code to the mechanism.
24 lessons, interactive models, Rust implementations and integrated exercises explain the internals.
Explore the implementation course →24 lessons, interactive models, Rust implementations and integrated exercises explain the internals.
Explore the implementation course →Change a short HTML input and follow the steps that turn a string into a tree.
Try the demo →OPEN TEXTBOOKS
Work through concrete examples and explore the mechanisms behind the code.
Algorithms and Data Structures
About 19 min
Computer Architecture
About 9 min
Computer Architecture
About 10 min
Computer Architecture
About 10 min
KNOWLEDGE MAP
Explore available lessons and upcoming topics. Select a topic to see its objectives and prerequisites.
Describe discrete structures using sets, relations, logic, proof, and induction.
Work with multidimensional structures through vectors, matrices, linear maps, eigenvalues, and decompositions.
Model uncertainty with random variables and distributions, then estimate and evaluate quantities from samples.
Connect rates of change, gradients, the chain rule, convexity, and optimization to machine learning and systems analysis.
Understand how programs execute, manage memory, express abstractions, and support debugging.
Study computation and data representation through correctness, complexity, and measured performance.
Understand how instructions, processors, pipelines, caches, I/O, and multiple cores execute programs.
Explore the limits of computation through formal languages, automata, computability, reductions, and complexity classes.
Understand resource abstraction and protection through virtualization, concurrency, and persistence.
Trace communication through layers, addressing, routing, reliability, name resolution, and web protocols.
Understand persistent data management through data models, indexes, queries, transactions, and recovery.
Reason about races, synchronization, ordering, and progress when executions overlap.
Reason about consistency and availability across nodes that can fail independently and communicate with delays.
Connect syntax, semantics, and types to program analysis, transformation, and runtime systems.
Understand resource abstraction and operations using virtual machines, hypervisors, containers, and orchestration.
Analyze threat models, authentication, authorization, memory safety, web attacks, isolation, and side channels.
Understand the primitives and assumptions behind confidentiality, integrity, authentication, and key exchange.
Protect execution and data from an untrusted host and verify execution state remotely.
Understand learning from data through loss minimization, generalization, evaluation, and representation.
Study neural networks, backpropagation, optimization, and major model architectures.
Understand language models through tokenization, self-attention, pretraining, generation, and KV caching.
Connect accelerator execution, inference serving, vector search, RAG, evaluation, observability, and security.