Performance Optimization¶

Tier: 3 (Applied Domains)

Concepts: 14

Durability: Timeless (rooted in physical constraints - finite resources, thermodynamics, mathematical limits on parallelization)

Fundamental Tensions¶

Tension	Description
Latency vs. Throughput	Optimizing for response time per request often conflicts with maximizing total work per time unit; batching improves throughput but increases individual latency
Memory vs. Computation (Space-Time)	Storing precomputed results saves CPU cycles but consumes memory; this trade-off is fundamental to caching, memoization, and lookup tables
Simplicity vs. Performance	Optimized code is often harder to read, maintain, and modify; clarity and speed frequently pull in opposite directions
Generality vs. Optimization	Generic solutions handle many cases adequately; specialized solutions handle specific cases excellently - optimization typically requires sacrificing generality
Correctness vs. Speed	Some optimizations (caching, denormalization, eventual consistency) introduce complexity that risks correctness; the fastest code that produces wrong answers is worthless

This Domain Concept	Transfers To
Measure Before Optimizing	Scientific method, quality improvement, operations research, business analytics
Amdahl's Law	Project management (critical path), manufacturing, queuing theory, parallel algorithms
Bottleneck-First Optimization	Manufacturing (TOC), supply chain, project scheduling, network design
Caching as Space-Time Trade-off	Hardware architecture, distributed systems, web infrastructure, compilers
Cache Invalidation Problem	Distributed consensus, data synchronization, derived data management
Latency vs. Throughput Trade-off	Manufacturing, logistics, network protocols, scheduling systems
Resource Pooling	Database systems, operating systems, network services, any resource management
Lazy Evaluation	Functional programming, OS memory management, query optimization
Memoization	Dynamic programming, build systems, API design, computational mathematics
Batching to Amortize Overhead	I/O systems, transaction processing, network protocols, batch computing
Precomputation	Database indexing, content delivery, game engines, scientific computing
Diminishing Returns	Economics, project estimation, resource allocation, optimization theory
Load Testing	Infrastructure engineering, capacity planning, release management
Compression Trade-off	Network systems, storage systems, media processing, data pipelines

Source	Coverage
Amdahl, "Validity of the Single Processor Approach to Achieving Large Scale Computing Capabilities" (1967)	Amdahl's Law, parallelization limits
Knuth, "Structured Programming with go to Statements" (1974)	Premature optimization, measurement methodology
Gregg, "Systems Performance: Enterprise and the Cloud" (2020)	Profiling methodology, bottleneck identification, comprehensive performance analysis
Goldratt, "The Goal" (1984)	Theory of Constraints, bottleneck-first thinking
Little, "A Proof for the Queuing Formula: L = lambda W" (1961)	Little's Law, queuing fundamentals
Gray & Reuter, "Transaction Processing: Concepts and Techniques" (1992)	Batching, resource management, throughput optimization
Hennessy & Patterson, "Computer Architecture: A Quantitative Approach"	Caching, memory hierarchy, performance modeling
Bentley, "Programming Pearls" (1986) & "Writing Efficient Programs" (1982)	Practical optimization techniques, profiling
Hughes, "Why Functional Programming Matters" (1989)	Lazy evaluation benefits
Michie, "Memo Functions and Machine Learning" (1968)	Original memoization concept
Shannon, "A Mathematical Theory of Communication" (1948)	Information theory foundations for compression
Ousterhout, "A Philosophy of Software Design" (2018)	Performance vs. clarity trade-offs