🐭 计算机废物自学指北

ETH Zurich Digital Design & Computer Architecture: SystemVerilog labs building a MIPS/RISC-V processor, verified in simulation

GAMES101 (Lingqi Yan) — all 8 assignments as CPU software renderers in C++/Eigen: rasterizer, MSAA, Blinn-Phong, Bezier, Whitted ray tracing, BVH, Cornell-box path tracing

Full solutions to UC Berkeley CS188 Pacman AI projects — search, adversarial games, RL, probabilistic inference, and ML, all passing the official autograders

MIT 6.007 Signals and Systems: from-scratch Python for LTI/convolution, Fourier series & transforms, sampling & aliasing, Laplace/z-transforms, filtering, modulation, and feedback — 64 pytest checks, 10 figures

Full-stack BadgerChat for UW-Madison CS571: a JWT REST API with React (web) + React Native (Expo) clients — 79 tests

Full Stack Open (University of Helsinki) core parts 3-8: a running Bloglist full-stack app (Express + Mongoose + JWT REST API, React + React Query + Router SPA) and a Library GraphQL API (Apollo Server v4), 52 tests + live E2E smoke checks.

CPU-only Python/NumPy reimplementation of CMU 15-462/662 Scotty3D: software rasterizer, half-edge MeshEdit (subdivision + QEM), BVH path tracer, and skeletal animation — real renders + measured results

HIT 哈工大《操作系统》(李治军) hit-oslab labs: real Linux 0.11 kernel modifications (bootloader, syscalls, scheduler trace, semaphores, shared memory, terminal, /proc) verified under Bochs. A csdiy.wiki build.

Solutions to UC Berkeley CS61B (Data Structures, Java) — including Gitlet (a Git version-control clone) and BYOW — with the official tests passing

From-scratch USTC Computer Graphics (Ligang Liu): image warping, Poisson editing, minimal surface, Tutte/ARAP parameterization, ARAP deformation, QEM simplification, software rasterizer & Monte-Carlo path tracer — CPU Python, real rendered results

UC Berkeley CS168 Introduction to the Internet — programming projects on reliable transport, routing, and traceroute

Caltech CS122 NanoDB (Java): storage & buffer management, B+tree indexing, SQL translation, query planning/optimization, and transactions

Rigorous solutions + verified experiments for Stanford STATS214/CS229M Machine Learning Theory (concentration, Rademacher complexity, margins, kernels, NTK, implicit regularization, online learning)

Working, verified solutions to a representative subset of Syracuse SEED Labs across software, crypto, network, and web security (WSL2, real captured evidence). Educational.

Solutions to MIT 6.100L Introduction to CS and Programming using Python — all problem sets and finger exercises, verified by running on the provided test cases

Solutions to Hung-yi Lee (李宏毅) Machine Learning course homeworks HW1-HW15 with measured results

Stanford CS242 Programming Languages: assignments on lambda calculus, type systems, and language implementation

Solutions to CMU 11-785 Introduction to Deep Learning — a from-scratch autograd library (mytorch: MLP, CNN, RNN/GRU + CTC, attention) plus the Kaggle deep-learning challenges — with the official autograder passing

Shanghai Jiao Tong University Compilers: a compiler implementation for a C-like language across the standard phases