Compute & Memory Foundations¶

Tier: 0 (Computational Foundations)

Concepts: 14

Durability: Timeless (rooted in physical constraints of silicon, speed of light, and mathematical limits of numeric representation)

Fundamental Tensions¶

Tension	Description
Speed vs. Capacity	Faster memory requires physical proximity to the processor and simpler circuitry, inherently limiting capacity. This creates the memory hierarchy - a fundamental trade-off that cannot be eliminated without violating physics (speed of light, signal propagation).
Locality vs. Abstraction	High-level abstractions (objects, closures, indirection) improve programmer productivity but scatter related data in memory, harming locality. Clean code and cache-friendly code often pull in opposite directions.
Precision vs. Performance	Exact arithmetic requires arbitrary precision and is computationally expensive. Fixed-size floating point enables hardware acceleration but introduces representation errors. This is a mathematical necessity - infinite real numbers cannot fit in finite bits.
Predictability vs. Flexibility	Processors optimize for predictable patterns (sequential access, consistent branches). Flexible, data-dependent code paths defeat these optimizations. Hardware speculation requires regularity to be effective.
Isolation vs. Efficiency	Virtual memory provides process isolation and address space abstraction, but adds indirection cost. Direct physical access would be faster but unsafe. Security and performance trade off at the memory management layer.

This Domain Concept	Transfers To
Memory Hierarchy	Database buffer pools, CDN architecture, distributed caching, storage tiering
Locality Principles	Query optimization, garbage collection, JIT compilation, recommendation systems
Cache-Conscious Design	Column stores, entity-component systems, scientific computing
Latency Numbers Intuition	Distributed systems, API design, capacity planning
CPU-bound vs I/O-bound	System design, scaling strategies, language selection
Branch Prediction	Query execution, virtual machines, compiler optimization
Instruction Pipelining	Stream processing, CI/CD design, assembly line optimization
SIMD/Data Parallelism	Machine learning, GPU computing, data processing
Floating-Point Limits	Financial systems, scientific computing, ML training
Floating-Point Pitfalls	Numerical algorithms, physics simulations
Integer Overflow	Security, embedded systems, financial calculations
Virtual Memory	Containerization, database design, runtime memory models
Page-Based Management	File systems, allocators, virtualization
Memory-Mapped I/O	Database buffer pools, IPC, high-performance I/O

Source	Coverage
Patterson & Hennessy, Computer Organization and Design (1994+)	Memory hierarchy, CPU architecture, pipelining, I/O
Hennessy & Patterson, Computer Architecture: A Quantitative Approach	Advanced CPU concepts, performance analysis
Drepper, "What Every Programmer Should Know About Memory" (2007)	Memory hierarchy, caching, NUMA, practical optimization
Goldberg, "What Every Computer Scientist Should Know About Floating-Point Arithmetic" (1991)	Floating-point representation, precision, arithmetic pitfalls
Denning, "The Working Set Model for Program Behavior" (1968)	Locality, virtual memory theory, working sets
Silberschatz et al., Operating System Concepts	Virtual memory, process management, memory protection
Dean & Barroso, "The Tail at Scale" (2013)	Latency at scale, percentiles, system design
Flynn, "Very high-speed computing systems" (1966)	SIMD classification, parallel architectures
Higham, Accuracy and Stability of Numerical Algorithms	Numerical analysis, floating-point stability
Fog, Optimizing software in C++ and microarchitecture manuals	Practical CPU optimization, pipeline details