Computer Science
Algorithms, data structures, computation theory, and systems design
Rowhammer: The DRAM Vulnerability That Flips Bits by Reading Memory
Rowhammer is a hardware vulnerability in DRAM where repeatedly accessing (hammering) one memory row causes electrical interference that flips bits in physically adjacent rows. Discovered in 2014 by Yoongu Kim et al. at CMU, it affects 85%+ of DRAM modules tested. Google Project Zero demonstrated it could be weaponized for privilege escalation — an unprivileged process flipping bits to gain kernel access.
Vector Databases: How Embedding Search Powers Modern AI Applications
Vector databases are specialized storage systems optimized for similarity search on high-dimensional embedding vectors. They enable semantic search — finding items by meaning rather than keywords — by storing numerical representations (embeddings) of text, images, or other data and quickly finding the nearest neighbors to a query vector. Key implementations include pgvector (PostgreSQL extension), Pinecone, Weaviate, Chroma, Qdrant, and Milvus. They are the retrieval backbone of RAG (Retrieval-Augmented Generation) systems.
Reversing Pseudo-Random Number Generators: Why PRNGs Are Predictable and Hackable
Most computer-generated 'random' numbers come from deterministic formulas that are fully reversible. Linear Congruential Generators can be inverted with modular multiplicative inverses. XOR-shift generators can be unwound step by step. JavaScript's Math.random(), Minecraft's world generation, PHP, and Flash all use hackable PRNGs. Only cryptographically secure RNGs (CSPRNGs) are safe for security-sensitive applications.
CPU Cache Levels Explained: Why L1 Is Faster Than L3
CPU cache speed differences come from physical distance to the core and size tradeoffs. L1 (~1ns) is tiny but inside the core. L3 (~10-40ns) is larger but farther away. At 3 GHz, even millimeters matter.
Why Computers Aren't More Modular: The Interconnect Bottleneck
Hardware modularity is limited by interconnect speeds (inter-chip communication is ~100x slower than intra-chip), software parallelization difficulty, and Amdahl's Law. Apple's integrated SoC approach avoids the bottleneck.
P vs NP Problem: Fundamentals Explained
P = efficiently solvable, NP = efficiently verifiable. NP-complete problems are the hardest — solving any one efficiently would solve all NP problems. Proof must be mathematical, not algorithmic.