Permutations & Combinations

You're seeing it when your feature configuration allows users to set 8 binary toggles, creating 256 possible states. The QA team can't test every combination — they need a structured approach to coverage.

You're seeing it when you're building a founding team and the order of hires matters: hiring an engineer first then a designer creates a different culture than the reverse. The permutations are distinct outcomes, not just different sequences.

You're seeing it when a go-to-market plan involves choosing 3 channels from 12 options. The 220 possible combinations make it clear that intuition alone can't evaluate the full space — you need selection criteria.

When facing multi-variable decisions — team composition, feature bundles, market entry sequence — estimate the combinatorial space first. If it's large, define constraints to reduce it before analyzing options. Don't waste cycles on exhaustive evaluation when structured elimination gets you to a decision faster.

Huang built NVIDIA into a computing powerhouse by understanding combinatorial scale. GPU architecture is fundamentally a combinatorics problem: how to process massive parallel computations across thousands of cores simultaneously. Huang recognized that the explosion of possible data paths and processing configurations in parallel computing required architectural choices that constrained the combinatorial space intelligently rather than brute-forcing every possibility. The same thinking applied to NVIDIA's strategic positioning: instead of chasing every possible market (the full combinatorial set), Huang selected the specific combination of gaming, data centers, and AI that created compounding advantages. Founders face equivalent choices — the combinatorial space of markets, products, and channels is vast, and the winners constrain it deliberately.