Assertion Partly supported AI assessment confidence: 90% certainty 4/5 debate potential 3/5

Goldberg: Moore's Law has slowed from 18 months to five years

Jay Goldberg · Why are chips interesting again? · Dec 19, 2022 · at 17:35

Venture partner Jay Goldberg discusses the deceleration of semiconductor manufacturing improvements with Benedict Evans and Tony Karen Brown.

0:00 / 0:04exact quote · 4.7s

720p mp4 · rendered on demand · StarZero watermark

“Moore's law is slowing, right? It is not 18 months anymore. It's five years.”

quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →

The claim, plainly

Moore's law cadence has slowed from 18 months to five years.

a quote alone can be ambiguous, so this restates it with the references and context of the conversation around it resolved. AI-written; the tape below the player is the source.

Assessment

The claim bundles two assertions: that Moore's law scaling has slowed beyond its historical 18- to 24-month cadence, and that the new cadence has reached five years. As of 2022–2023, empirical data from leading foundries (TSMC, Samsung, Intel) and industry benchmarks confirm that transistor density scaling has indeed slowed, but the doubling cadence lengthened to approximately 2.5 to 3.5 years rather than five years. For instance, TSMC increased logic density by ~3.2x between N7 (April 2018) and N3 (December 2022), yielding an effective doubling period of ~2.8 years, while Intel CEO Pat Gelsinger noted in 2022–2023 that doubling was occurring closer to every three years.

TSMC progressed from N7 (April 2018, ~91.2 MTr/mm²) to N3 (December 2022, ~290 MTr/mm²), achieving a ~3.2x density increase in 4.7 years (an effective doubling time of ~2.8 years).
Intel CEO Pat Gelsinger publicly stated in 2022–2023 that transistor doubling had slowed to 'closer to every three years now' rather than five years.
Industry consensus and benchmark roadmaps (IRDS, SemiWiki, TechInsights) indicated leading-edge process node doubling cycles stretched to between 2.5 and 3.5 years by 2022.

Checked against

[1] wikipedia.org [2] reddit.com [3] wikipedia.org [4] wordpress.com [5] angstronomics.com [6] wikipedia.org [7] 36kr.com [8] ycombinator.com

"Moore's Law" "cadence" "five years" 2022 OR 2023 transistor density MTr mm2 TSMC Samsung Intel nodes timeline "transistor density" "doubling time" "years" TSMC 2022 OR 2023 TSMC N7 N5 N3 transistor density scaling factor timeline transistor density scaling TSMC Intel Samsung cadence 2018 2020 2022 2023 "Pat Gelsinger" "three years" "Moore's law" 2022 OR 2023 TSMC transistor density doubling time Moore's law cadence 2022 Moore's law cadence slowed TSMC N7 N5 N3 transistor density doubling

assessed as of 2026-09-02

This assessment is an AI reading of a public statement, made against the cited references with the confidence shown. It can be wrong. The quote, the tape and the references are all on this page; if something is off, use report an error below. full disclaimer →

Mentioned

Moore's law

⚑ Spot an error? Report it

More from Jay Goldberg

Opinion

Goldberg: Only 'one and a half' foundries can produce leading-edge chips

“There's really only one and a half that can really produce at the leading edge, right? TSMC in Taiwan is just way ahead of everybody else. Samsung is kind of the half. They're close, but not quite there. And then there's nobody else.”

Jay Goldberg Dec 19, 2022 ▶ 5:03 Why are chips interesting again?

Opinion

Goldberg: Qualcomm's Snapdragon processors trail Apple silicon by two years

“And so you look at the current Snapdragon applications processes that Qualcomm makes, it's, I don't know, it's two years behind where Apple's A series is or N series is.”

Jay Goldberg Dec 19, 2022 ▶ 8:43 Why are chips interesting again?

Assertion Supported

Goldberg: Cost per transistor is increasing in many cases

“The cost per transistor is actually going up in a lot of cases”

Jay Goldberg Dec 19, 2022 ▶ 18:33 Why are chips interesting again?

Prediction Partly held up

Goldberg: TSMC's Arizona fab will lag leading-edge chips by half a generation

“And now I think that's actually gonna be a real plant in Arizona. And I think it's gonna have decent capacity enough to sort of satisfy the U.S. Military's needs and Apple's needs. And it will be, I don't think it'll be even a generation behind. I think it'll …”

Jay Goldberg Dec 19, 2022 ▶ 25:11 Why are chips interesting again?

Prediction Not checkable as stated

Goldberg: TSMC's Arizona fab will remain dependent on Taiwan's expertise

“It doesn't mean the U.S. Actually can do advanced chips because that plant is going to be entirely dependent on expertise and knowledge base from the mothership in Taiwan.”

Jay Goldberg Dec 19, 2022 ▶ 26:02 Why are chips interesting again?

Assertion Not checkable as stated

Goldberg: China has failed to copy TSMC by poaching personnel

“And China for years has been trying to poach people out of TSMC to come build their foundries. And it just hasn't worked. Like you, it's something about the whole body of knowledge that's just too, you just can't copy and paste.”

Jay Goldberg Dec 19, 2022 ▶ 27:01 Why are chips interesting again?

Goldberg: Moore's Law has slowed from 18 months to five years

More from Jay Goldberg

Turn any episode into a week of clips.