what is moores law
Moore's Law is the observation, made by Intel co-founder Gordon Moore in 1965, that the number of transistors that can be economically packed onto a chip roughly doubles every couple of years — and for over half a century, that steady doubling has been the underlying engine behind computers, phones, and every other chip-powered device getting dramatically more capable at roughly the same price.
```flowchart
{
"rows": [
{ "type": "nodes", "items": [
{ "title": "1965: Gordon Moore's observation", "sub": "transistor count on a chip doubles roughly every 2 years", "tone": "neutral" }
]},
{ "type": "arrow" },
{ "type": "group", "title": "Decades of compounding doubling", "items": [
{ "title": "More transistors per chip", "sub": "smaller, more numerous, more capable", "tone": "green" },
{ "title": "Lower cost per transistor", "sub": "same budget buys far more computing", "tone": "green" }
]},
{ "type": "arrow" },
{ "type": "nodes", "items": [
{ "title": "Today: physical and economic limits slow the pace", "sub": "shrinking further gets harder every generation", "tone": "orange" }
]}
]
}
```
**Moore's Law was never a law of physics — it was an economic and engineering observation that became a self-fulfilling industry roadmap.** Gordon Moore noticed a trend in the early years of the chip industry and projected it forward; because the observation proved broadly accurate for so long, chipmakers, equipment suppliers, and researchers all planned their roadmaps around continuing to hit that pace, turning the prediction into an industry-wide target that companies actively worked to meet rather than a fixed natural rule they simply happened to follow.
```svg
```
| Era | What was driving the doubling | Typical challenge |
|---|---|---|
| 1970s-1990s | Straightforward shrinking of transistor dimensions | Manufacturing precision at ever-smaller scales |
| 2000s | Shrinking plus new materials and transistor structures | Power and heat from ever-denser chips |
| 2010s-today | New transistor shapes, 3D structures, advanced packaging | Physical limits nearing atomic scale, rising cost per generation |
**The pace of Moore's Law has clearly slowed compared to its fastest decades, and the industry is largely open about why.** Transistors are now only a handful of atoms across in some dimensions, so squeezing them smaller runs into genuine physics limits, not just engineering difficulty; at the same time, each new generation of manufacturing equipment and process development costs dramatically more than the last, meaning the economic half of "smaller and cheaper" has gotten much harder to sustain even where the physics still technically allows further shrinking.
**The chip industry has responded by finding gains outside of simply shrinking transistors further, sometimes summarized as "More than Moore."** Advanced packaging techniques that stack multiple chips together, chiplet designs that combine several smaller dies into one package, and specialized chips built for one task instead of general-purpose computing have all become major sources of continued performance improvement — meaning the historical trend of "computers get meaningfully better every couple of years" has largely continued, even as its original mechanism has changed.
**Moore's Law is often confused with a related but separate idea about performance, which causes a lot of misunderstanding about what actually slowed down.** Transistor count doubling doesn't automatically mean a chip runs twice as fast in every task — it means twice as many switches are available to build with, which can go toward more cores, larger caches, specialized processing blocks, or genuinely faster individual transistors, depending on what a given chip design prioritizes. A generation with a modest raw speed increase can still represent a full, on-schedule Moore's Law doubling if that transistor budget went mostly toward new capabilities instead of pure clock speed.
Read Moore's Law through a trend-turned-roadmap lens: it began as an observation about a young industry's early progress, but became a coordinating target the entire chip industry organized itself around for decades — and understanding today's slower pace means recognizing that both the physics and the economics behind that original observation have genuinely changed, not that progress itself has stopped.