How many transistors fit on a chip?
Transistors per microprocessor — Moore's law · count · measured 1971–2021 · fitted live from 39 observations by the Many Minded cost-curve engine.
The last measured value is 58.2B (2021, OWID / Karl Rupp). Over the fitted window the series rises 34.1% a year (80% interval +17.6% to +52.9%) — a doubling every 2.4 years (80%: 1.6–4.3). Carried forward on that fit, 2026: 252B (80%: 78.2B–815B).
Last measured 2021. The source has not published a newer figure we have verified, so the 5 years since are projections too, not observations.
the curve
what this series measures
transistor count of flagship microprocessors — the Moore’s-law capability climb.
what the fit says
| quantity | value | how it is computed |
|---|---|---|
| fitted rate | +34.1%/yr (80%: +17.6% to +52.9%) | Farmer–Lafond drift over the trailing window · n=11 points, 11 years |
| doubling time | 2.4 years (80%: 1.6–4.3) | implied by the fitted drift |
| curve check | accelerating (40% → 43%/yr) | first half of the record versus the second, judged against the direction that helps this metric (rising) |
| regime break | ~2002 (39% → 36%/yr, F=20.8) | Chow-style best single breakpoint, kept only above a conservative sup-F threshold |
Not shown for this curve, because the machinery returns nothing: Wright's law (no cumulative-deployment series for this technology).
the projection, with its interval
Log cost as a random walk with drift: the forecast variance grows with horizon (τ + τ²/m), which is why these bands widen instead of staying parallel. The middle column is the least useful number on this page; the interval is the claim.
| year | central fit | 80% interval | |
|---|---|---|---|
| 2022 | 78.0B | 49.6B to 123B | near horizon |
| 2023 | 105B | 53.7B to 204B | near horizon |
| 2024 | 140B | 60.0B to 328B | near horizon |
| 2025 | 188B | 68.2B to 519B | the band is already wide here |
| 2026 | 252B | 78.2B to 815B | the band is already wide here |
| 2027 | 338B | 90.1B to 1.27T | the band is already wide here |
| 2028 | 454B | 104B to 1.97T | far horizon — read the interval, not the middle |
| 2029 | above 582B | — | past the point where a number would be theater |
The table stops at 582B — a decade above the highest value ever observed here. Past that, quoting a number would be theater rather than forecast.
questions this page answers
How many transistors fit on a chip?
58.2B as of 2021, the latest measured value in the series (OWID / Karl Rupp). The fitted trend has it rising 34.1% a year, with an 80% interval of +17.6% to +52.9%.
How fast is the transistor count of a flagship chip rising?
+34.1% a year over the fitted window, an 80% interval of +17.6% to +52.9% — a doubling every 2.4 years (80%: 1.6 to 4.3 years). Fitted from 11 observations spanning 11 years.
What will the transistor count of a flagship chip be in 2026?
The central fit says 252B, inside an 80% interval of 78.2B to 815B. The interval is the forecast; the middle number is only its midpoint. Bands widen with horizon because shocks accumulate — a constant-width band would be overconfident.
Is the rise in the transistor count of a flagship chip accelerating or slowing?
Splitting the record in half, the fitted rate went from 40% to 43% a year — accelerating. A Chow-style test finds a regime break around 2002 (39% → 36% a year, F=20.8). Regime breaks, not window choice, are what dominate this method's errors — which is why every projection here carries an interval.
Where does this data come from?
OWID / Karl Rupp. 39 observations spanning 1971–2021. Curated benchmark history, extended by a weekly authoritative fetch and by news figures fact-checked against their source before they may touch a fit. Both the observation ledger and the fitting code are public, and the engine publishes its own calibration score and its misses.
the raw numbers
Every observation behind the fit, unrounded by us and unsmoothed. This table is here on purpose: graphs make people underestimate exponential change, and the raw series beside the curve is the one correction shown to work.
| year | count | change |
|---|---|---|
| 1971 | 2.31K | — |
| 1972 | 3.56K | +54.0% |
| 1974 | 6.10K | +31.0%/yr |
| 1979 | 29.2K | +36.7%/yr |
| 1982 | 136K | +67.0%/yr |
| 1985 | 274K | +26.3%/yr |
| 1986 | 274K | +0.0% |
| 1988 | 274K | +0.0%/yr |
| 1989 | 1.21M | +341.2% |
| 1990 | 1.21M | +0.0% |
| 1992 | 3.11M | +60.3%/yr |
| 1993 | 3.11M | +0.0% |
| 1994 | 3.11M | +0.0% |
| 1995 | 9.65M | +210.6% |
| 1996 | 9.65M | +0.0% |
| 1997 | 9.65M | +0.0% |
| 1998 | 15.3M | +58.2% |
| 1999 | 21.7M | +42.0% |
| 2000 | 37.2M | +71.6% |
| 2001 | 42.5M | +14.4% |
| 2002 | 221M | +418.7% |
| 2003 | 221M | +0.0% |
| 2004 | 274M | +24.1% |
| 2005 | 305M | +11.4% |
| 2006 | 583M | +91.1% |
| 2007 | 806M | +38.2% |
| 2008 | 806M | +0.0% |
| 2009 | 2.31B | +186.4% |
| 2010 | 2.31B | +0.0% |
| 2011 | 2.60B | +12.7% |
| 2012 | 2.60B | +0.0% |
| 2013 | 5.00B | +92.3% |
| 2014 | 5.70B | +14.0% |
| 2016 | 8.00B | +18.5%/yr |
| 2017 | 19.2B | +140.0% |
| 2018 | 21.1B | +9.9% |
| 2019 | 39.5B | +87.2% |
| 2020 | 39.5B | +0.0% |
| 2021 | 58.2B | +47.3% |
where this comes from
- Source: OWID / Karl Rupp — refetched weekly by the engine's benchmark cron.
- Basis: measured. Curated benchmark history, refreshed by the weekly authoritative fetch, extended by figures our daily editor extracts from the news — each of which is fact-checked against its source article before it may touch a fit.
- Observations: n=39, spanning 1971–2021. Ledger last updated 2026-11-02.
- Fit: the Farmer–Lafond stochastic model — log cost as a random walk with drift over a 10-year trailing window, Student-t tails, and (for series under 15 points) a volatility floor at their drift-volatility prior. The code is readable source; the observation ledger is public JSON.
- Reuse: this compiled series and the numbers on this page are published under CC BY 4.0 — take them, cite the page. The underlying source keeps its own terms (Our World in Data publishes under CC BY 4.0).
- Accuracy: the engine grades its own intervals in public — every feasible hindcast re-fitted on data as it stood, scored for whether the actual landed inside the 80% band. The calibration score is on the engine page, misses included.
the rest of the graph
Other KNOWLEDGE curves:
- Hard drive cost per terabyte $/TB
- Fastest supercomputer, computational capacity GFLOP/s
- DRAM memory cost per terabyte $/TB
- Solid-state storage cost per terabyte $/TB
- AI task horizon at 50% reliability (METR) minutes
- Frontier AI training compute, FLOP FLOP
- Frontier-grade AI inference price $/Mtok
Recent briefs from KNOWLEDGE, the daily record of what moved:
- A detector that reads a paper's argument structure to spot AI authorship 2026-11-08University of Maryland researchers built IdeaLens, which flags AI-written text by the shape of its ideas, not its wording. It is still in development.
- Ripple-65B Preview: Open-Source Coding Model Shows Promise But Needs Refinement 2026-11-07Ricursive Intelligence released a 65-billion-parameter coding model with strong benchmark results, but it remains a preview model not yet ready for production use.
- AI research agents still fail to meet expert standards in autonomous scientific work 2026-11-07Nine AI agents tested on open research questions from US national labs failed to meet experts' standards for credible scientific output
- Schools get Gemini switched on first, and a real off-switch only after pushback 2026-11-07Google enabled Gemini for 150 million Classroom students, then split its controls after criticism. EFF argues the default still favors Google, not schools.
- Chinese open-weight models become a default pick for enterprises going open 2026-11-06Low-cost open-weight models from Chinese labs, Moonshot AI's Kimi among them, are reportedly leading enterprise open-source AI spending.
- X-ray evidence points to altermagnetism in manganese telluride 2026-11-06A SLAC-led team reports direct evidence of handedness in manganese telluride's magnetic structure, supporting a third class of magnetism.
Every tracked curve: utility solar, installed · battery pack price · battery cell price · solar module price · human genome sequencing · onshore wind electricity · solar electricity (LCOE) · space launch, to LEO · wheat yield · offshore wind electricity · modular housing, turnkey · AI video, finished minute · humanoid robot, capable unit · cheapest humanoid, list price · people still offline · people outside a mobile-broadband signal · farm labor share · healthcare spend per person
← the whole engine: build clock, cascade, calibration · the archive · how we know