But what I realized is actually the exact opposite.

但我意識到的情況其實恰恰相反。

If you take nuclear weapons, that technology has barely improved since the 1960s.

就拿核武器來說，自 20 世紀 60 年代以來，該技術幾乎沒有任何改進。

It's so easy to make nuclear bombs, even the North Koreans can do it.

製造核彈太容易了，連朝鮮人都能做到。

But chips are everywhere because they're cheap and they're tiny.

但薯片卻隨處可見，因為它們既便宜又小巧。

And making things very inexpensive and very small is extraordinarily difficult.

而要把東西做得既便宜又小巧，難度可想而知。

Today, there are just three companies capable of producing cutting-edge processor chips, the types of chips that go in phones or computers or are used for AI.

如今，只有三家公司有能力生產最先進的處理器芯片，即用於手機、電腦或人工智能的芯片。

For many of those devices, they rely on chips that, in some cases, can only be made by one company in a single factory in Taiwan.

其中許多設備都依賴於芯片，而在某些情況下，這些芯片只能由一家公司在臺灣的一家工廠生產。

The key risk that hangs over not just the chip industry, but really our entire economy, is that something goes wrong between China and Taiwan and the Taiwan Straits, such that they lose access to chips made in Taiwan.

不僅是芯片產業，實際上我們整個經濟都面臨的主要風險是，中國大陸和臺灣以及臺灣海峽之間出現問題，導致他們無法獲得臺灣製造的芯片。

My name is Chris Miller.

我叫克里斯-米勒。

I'm a professor at the Fletcher School and author of Chip War, the fight for the world's most critical technology.

我是弗萊徹學院的教授，也是《芯片戰爭：爭奪世界最關鍵技術》一書的作者。

Well, I first got interested in chips when I realized you really couldn't understand how the world works without them.

我第一次對芯片感興趣，是因為我意識到，沒有芯片，就無法理解世界是如何運轉的。

When we think about technology, we think about social media, we think about search engines, we think about apps on our phones, but undergirding all this are chips.

提到科技，我們會想到社交媒體、搜索引擎、手機上的應用程序，但支撐這一切的是芯片。

A chip is a piece of silicon, often the size of your fingernail, and in it is carved thousands or millions, in some cases, billions of tiny devices called transistors, which flip circuits on or off, on and off.

芯片是一塊硅片，通常只有指甲蓋大小，上面雕刻著數千或數百萬個，有時甚至數十億個被稱為半導體的微小器件，這些器件可以接通或斷開電路。

And when they're on, they produce a one.

當它們啟動時，就會產生一個。

When they're off, they produce a zero.

當它們關閉時，就會產生一個零。

And all of the ones and zeros, undergirding computing, undergirding data storage, all of your Instagram likes, all of your text messages, these are all just long strings of ones and zeros, which are created on the chip by these circuits flipping on and off.

所有的 "1 "和 "0"，支撐著計算，支撐著數據存儲，支撐著你在 Instagram 上的所有點贊，支撐著你的所有簡訊，這些都只是一長串的 "1 "和 "0"，通過這些電路的開關在芯片上產生。

Before transistors, computers used vacuum tubes, which were sort of light bulb-like devices that would turn on and off, on and off to produce the ones and zeros.

在半導體出現之前，計算機使用的是真空管，一種類似燈泡的裝置，通過開、關、開、關來產生 1 和 0。

They were cutting edge for their time, but they had huge inefficiencies.

它們在當時是最先進的，但效率極低。

They wasted a lot of heat, for example.

例如，他們浪費了大量的熱量。

They worked pretty slowly.

他們工作得很慢。

And they also, because they created light, attracted moths.

同時，由於它們能產生光線，也吸引了飛蛾。

And so computers had to be regularly debugged in the early days of computing, which meant removing moths from the lights that they were attracted to.

是以，在計算機發展的早期，必須定期對計算機進行調試，這就意味著要把飛蛾從它們所吸引的燈光上移開。

In the middle of the 20th century, William Shockley, John Bardeen, and Walter Brattain invented the first transistor while they were working at Bell Labs.

20 世紀中葉，威廉-肖克利、約翰-巴丁和沃爾特-布拉坦在貝爾實驗室工作時發明了第一個半導體。

It was part of AT&T at the time.

當時它是 AT&T 的一部分。

Invented at the laboratories, the transistor brought Bell scientists one of the two Nobel Prizes they have won for discoveries in physics.

在實驗室發明的半導體為貝爾科學家帶來了他們因物理學發現而獲得的兩個諾貝爾獎中的一個。

They were initially planning to use these as part of the telephone network.

他們最初計劃將這些設備用作電話網絡的一部分。

Though it was primarily developed for the betterment of telephone communications, the versatile transistor now appears in a staggering variety of devices and equipment.

雖然半導體的開發主要是為了改善電話通信，但現在它已廣泛應用於各種裝置和設備中。

Individual transistors were connected via wires in a way that was okay if you had a handful of transistors, but if you had a thousand connected together, you had a jungle of wires you had to manage.

單個半導體通過電線連接，如果只有幾個半導體還可以，但如果有上千個半導體連接在一起，就會出現需要管理的電線叢林。

The first chips were invented by engineers working at Texas Instruments and a company called Fairchild Semiconductor in Silicon Valley.

第一批芯片是由德州儀器公司和硅谷一家名為飛兆半導體公司的工程師發明的。

The first engineers realized that you could take multiple transistors and make them on a single piece of semiconductor material.

第一批工程師意識到，可以在一塊半導體材料上製造多個半導體。

And so that was the first chip, a piece of material with multiple transistors carved into it.

這就是第一塊芯片，一塊刻有多個半導體的材料。

This single silicon crystal will be trimmed, then sliced and diced.

這塊單晶硅晶體將被修剪，然後切片、切丁。

The objective, semiconductors.

目標是半導體。

Patterns are made by a photographic process and then micro-machined by chemical etching to produce nearly 1,700 transistors on each slice.

圖案是通過攝影工藝製作的，然後通過化學蝕刻法進行微加工，從而在每個切片上生產出近 1700 個半導體。

And so the jungle of connections was replaced by a single block of material, which was much more reliable and also much more easy to shrink in its size.

是以，連接的叢林被單一的材料塊所取代，後者更加可靠，也更容易縮小尺寸。

At first they were building chips primarily for the U.S. government, for the space program, for example, and for weapons systems.

起初，他們主要為美國政府製造芯片，例如用於太空計劃和武器系統。

But they realized early on you could take the exact same chips that the government wanted to guide spacecraft and use them for commercial applications like computers or pocket calculators.

但他們很早就意識到，你可以把政府希望用於引導航天器的芯片用於計算機或袖珍計算器等商業應用。

And that set the industry off into its first phase of growth in the 1960s and 70s.

這使得該行業在二十世紀六七十年代進入了第一個發展階段。

All the calculator know-how you'll probably ever need.

您可能需要的所有計算器知識。

Over time, a new set of companies emerged.

隨著時間的推移，出現了一批新的公司。

Intel, for example, was founded in 1969, and it quickly focused on making chips for personal computers, which at the time was a very small market.

例如，英特爾公司成立於 1969 年，很快就把重點放在為個人電腦製造芯片上，而這在當時還是一個很小的市場。

But they correctly bet that soon everyone would have a personal computer.

但他們正確地預測到，很快每個人都會擁有一臺個人電腦。

And Intel, even today, is the world's largest producer of chips that go inside PCs.

時至今日，英特爾仍是全球最大的個人電腦芯片生產商。

Gordon Moore was one of the two co-founders of Intel.

戈登-摩爾是英特爾公司的兩位創始人之一。

He's most famous today probably for coining the term Moore's law.

如今，他最有名的可能就是創造了摩爾定律這個詞。

Moore's law is not a law of nature.

摩爾定律不是自然法則。

It's not a law of physics.

這不是物理定律。

It's really a law of economics.

這其實是一條經濟規律。

Moore's law predicts that the number of transistors per chip and, as a result, the computing power per chip will double every couple of years.

摩爾定律預測，每塊芯片上的半導體數量以及每塊芯片的計算能力將每隔幾年翻一番。

And that's been empirically true since the 1960s.

自 20 世紀 60 年代以來，經驗證明了這一點。

If you're able to find a way to shrink, shrink your transistors smaller, you will be able to find a larger market as well.

如果你能找到縮小半導體的方法，你就能找到更大的市場。

And that has incentivized huge investments in shrinking, in improving manufacturing processes, and making chemicals more purified, which means that the capabilities of chips have gotten vastly better and continue to get much, much better at a faster rate than anything else.

這就促使我們在縮小體積、改進製造工藝、提高化學品純度等方面進行大量投資，這意味著芯片的性能已經大大提高，並將繼續以比其他任何東西都快的速度大大提高。

So I like to think, for example, of airplanes to illustrate the difference.

是以，我喜歡以飛機為例，來說明兩者的區別。

If airplanes doubled in speed every two years from the 1960s up to the present, we'd be flying faster, literally, than the speed of light.

如果從 20 世紀 60 年代到現在，飛機的速度每兩年翻一番，那麼我們的飛行速度就會超過光速。

But chips have done that.

但芯片已經做到了這一點。

Chips have increased in that capability because the scale of the transistors has shrunk.

由於半導體的尺寸縮小，芯片的功能也隨之增強。

Chips today are measured in nanometers.

如今的芯片以納米為組織、部門。

And so that makes them only slightly larger than atoms, far smaller than bacteria, smaller than a mitochondria, half the size for the most cutting-edge transistors of a coronavirus.

是以，它們只比原子稍大，比細菌小得多，比線粒體小，是冠狀病毒最尖端半導體大小的一半。

There's basically nothing we manufacture at such tiny scale.

我們基本上沒有生產過如此小規模的產品。

When you go inside one of these massive facilities called fabs, what you find is that there are huge machines and not much else, because humans are way too imprecise for manufacturing at nanometer scale.

當你走進這些被稱為 "晶圓廠 "的大型設施時，你會發現裡面只有巨大的機器，沒有其他東西，因為人類對於納米尺度的製造來說太不精確了。

The machines that make chips can cost $350 million a piece.

製造芯片的機器單價可達 3.5 億美元。

And they cost so much because they require some of the most precise components ever used, like a mirror that's the flattest mirror humans have ever made, a laser that's the most powerful laser ever deployed in a commercial device, and a ball of tin that falls through a vacuum and is struck twice by that laser, explodes into a plasma measuring 40 times the temperature of the surface of the sun.

它們之所以如此昂貴，是因為它們需要使用一些有史以來最精密的部件，比如一面人類製造的最平的鏡子，一臺有史以來在商業設備中使用的最強大的激光器，以及一個錫球，錫球穿過真空，被脈衝光擊中兩次後，爆炸成等離子體，其溫度是太陽表面溫度的40倍。

And this plasma emits light at just the right wavelength, 13.5 nanometers, to be bounced off the mirrors in exactly the right geometry and land on your chip to carve the transistors into the silicon.

這種等離子體發出的光波長恰好為 13.5 納米，能以正確的幾何形狀被反射鏡反彈到芯片上，從而在硅片上雕刻出半導體。

It's the most complex and expensive machine that humans have ever made.

這是人類有史以來製造的最複雜、最昂貴的機器。

And it's required to make all the most advanced chips.

所有最先進的芯片都需要它。

And that has enabled the explosion of computing power, both in terms of the computing capabilities in high power data centers or in your phone, but also the application of computing to all sorts of devices.

這使得計算能力爆炸式增長，既包括大功率數據中心或手機中的計算能力，也包括計算在各種設備中的應用。

Today, there's computing everywhere.

如今，計算無處不在。

It's in your dishwasher, it's in your refrigerator, it's in your coffee maker, it's in your car.

洗碗機裡有，冰箱裡有，咖啡機裡有，車裡也有。

And it's possible to put computing everywhere because today it's so cheap, we can produce it almost for free.

而且，我們有可能將計算技術應用到各個地方，因為如今的計算技術非常便宜，我們幾乎可以免費生產。

The chip industry was a global industry from really the earliest days.

芯片產業從一開始就是一個全球性產業。

Because chipmaking requires ultra purified materials and hugely complex equipment, everyone requires a set of partnerships to give them the materials and the intellectual property and the software and the tools that they need to produce advanced chips.

由於芯片製造需要超純材料和極其複雜的設備，是以每個人都需要建立一套夥伴關係，為他們提供生產先進芯片所需的材料、知識產權、軟件和工具。

So in the U.S. right now, most of the key chip firms only design chips.

是以，目前在美國，大多數主要芯片公司都只設計芯片。

Most of the manufacturing of chips happens in East Asia, in Taiwan, for example, or in Korea.

大部分芯片製造都在東亞，例如臺灣或韓國。

Many of the chemicals that go into chipmaking come from Japan.

許多用於芯片製造的化學物質都來自日本。

And the machines that are used to make chips come from either Silicon Valley, where some of them are still made, or the Netherlands or Japan.

用於製造芯片的機器要麼來自硅谷（其中一些仍在硅谷製造），要麼來自荷蘭或日本。

So the industry has globalized, but it's also specialized in the process.

是以，該行業已經全球化，但在此過程中也實現了專業化。

And so there's not a single region today that can make cutting edge chips on its own.

是以，如今沒有一個地區能夠獨立製造最先進的芯片。

If you take, for example, the primary processor inside of your smartphone, it was probably made in Taiwan, but it was made in Taiwan using chipmaking tools from the Netherlands and from the United States and from Japan.

以智能手機內部的主處理器為例，它很可能是在臺灣製造的，但在臺灣製造時使用了來自荷蘭、美國和日本的芯片製造工具。

It was produced using chemicals from Japan, and then often assembled and packaged in Malaysia before ending up inside of your smartphone.

它使用來自日本的化學品生產，然後通常在馬來西亞組裝和包裝，最後才裝進你的智能手機。

And that's typical.

這就是典型的例子。

During the pandemic, the supply and demand dynamics from the chip industry were out of whack.

在大流行病期間，芯片行業的供需動態失衡。

A lot of people ordered new computers, for example, to work from home.

例如，很多人訂購了新電腦，以便在家工作。

And so PC production shot up in ways that weren't expected, or people bought fewer cars in the early days of the pandemic.

是以，個人電腦的產量出現了意想不到的增長，或者在大流行病初期，人們購買汽車的數量減少了。

And so car production declined and companies couldn't predict what type of chip they would need.

是以，汽車產量下降，公司無法預測需要哪種類型的芯片。

The effect of that was to create shortages of certain types of chips.

其結果是造成了某些類型芯片的短缺。

Car companies in particular found they couldn't get the types of chips that they rely on.

特別是汽車公司發現，他們無法獲得他們所依賴的芯片類型。

The thing about cars is if you're missing just one chip, your car often doesn't work.

汽車的特點是，只要少了一個芯片，汽車往往就無法工作。

And during the pandemic, car companies found themselves often in that situation.

在大流行病期間，汽車公司發現自己經常處於這種境地。

Just a single chip, often even the cheapest chips, were causing them to have to leave cars in the factory parking lot as they waited for the right chip to arrive.

僅僅一個芯片，甚至往往是最便宜的芯片，就導致他們不得不把汽車停在工廠停車場，等待合適的芯片到來。

That created hundreds of billions of dollars of losses for manufacturers like automakers.

這給汽車製造商等製造商造成了數千億美元的損失。

And that matters because the shortages we saw in 2021 and 2022 are tiny in comparison to the shortages we would see if to a large-scale chipmaker like those in Taiwan.

這很重要，因為我們在 2021 年和 2022 年看到的短缺與臺灣等大規模芯片製造商看到的短缺相比微不足道。

The biggest chipmaker in the world is the Taiwan Semiconductor Manufacturing Company.

世界上最大的芯片製造商是臺灣半導體制造公司。

When it comes to advanced processor chips, like the chips in your phone or the chips in your computer, TSMC makes around 90% of them.

說到先進的處理器芯片，如手機中的芯片或電腦中的芯片，臺積電製造了其中的約 90%。

In 1987, there was an American engineer named Morris Chang.

1987 年，有一位名叫莫里斯-張的美國工程師。

The Taiwanese approached him and said, would you like to build a chip factory in Taiwan?

臺灣人找到他說，你願意在臺灣建立芯片工廠嗎？

And he said, yes.

他說，是的。

At the time, most chips were manufactured and designed by the same companies.

當時，大多數芯片都是由同一家公司製造和設計的。

But he established TSMC in Taiwan with the aim never of designing chips, only of manufacturing.

但他在臺灣成立臺積電的目的從來不是設計芯片，而是製造芯片。

That's exactly what TSMC has done.

臺積電正是這樣做的。

And it's enabled TSMC to win among its customers some of the largest companies in the world.

這也使臺積電贏得了一些全球最大公司的青睞。

Apple, NVIDIA, Qualcomm, AMD, they all rely on TSMC to produce its chips, which means that TSMC is the largest chipmaker in the world by far.

蘋果、英偉達、高通、AMD 都依靠臺積電生產芯片，這意味著臺積電是迄今為止全球最大的芯片製造商。

They've got an extraordinary market share and are arguably the most important company in the world because the chips that they produce, we rely on for basically everything.

他們的市場佔有率非常高，可以說是世界上最重要的公司，因為他們生產的芯片基本上是我們的一切。

China and Taiwan have been at odds ever since Chiang Kai-shek and what was left of the Nationalist Army fled the mainland for the islands back in 1949.

自 1949 年蔣介石和國民黨軍隊殘部逃離大陸前往臺灣島以來，中國和臺灣就一直不和。

Anything that disrupted chip production in Taiwan would be catastrophic for the world economy, especially if China carries through on the threats it regularly makes to use force against Taiwan to take control of the island.

任何破壞臺灣芯片生產的行為都將對世界經濟造成災難性影響，尤其是如果中國兌現其經常發出的威脅，對臺灣動武以控制該島的話。

China fired 11 ballistic missiles right over the island and encircled it with warships to prove it can strangle Taiwan whenever it wants.

中國向臺灣島上空發射了 11 枚彈道導彈，並用軍艦包圍了臺灣島，以證明它可以隨時扼殺臺灣。

Even a small move, a small conflict would be disastrous for the chip industry because Taiwan needs to import energy, needs to import chemicals, materials, tools, spare parts from Japan, from the United States, from Europe, energy coming in from the Middle East.

因為臺灣需要進口能源，需要從日本、美國、歐洲進口化學品、材料、工具和零配件，還需要從中東進口能源。

And if any of this is disrupted, chip production could break down.

如果其中任何一個環節受到干擾，芯片生產就會崩潰。

And if chip production in Taiwan breaks down, that matters for everyone because everyone uses Taiwanese-made chips.

如果臺灣的芯片生產出現問題，這對每個人都很重要，因為每個人都在使用臺灣製造的芯片。

Both China and the U.S. see chips as really central to the technology competition between them right now.

中美兩國都認為芯片是目前兩國技術競爭的核心。

China's worried that because it relies on importing chips from Taiwan and from Korea, which are both U.S. allies, it's going to be cut off in the future from getting the chips that it needs.

中國擔心，由於它依賴於從臺灣和韓國進口芯片，而這兩個地方都是美國的盟國，未來它將無法獲得所需的芯片。

And right now that's already happening to some degree.

而現在，這種情況已經在一定程度上發生了。

The U.S. is limiting the ability of AI firms like NVIDIA to sell their most cutting-edge chips to China.

美國正在限制英偉達（NVIDIA）等人工智能公司向中國出售其最尖端芯片的能力。

And the aim of these regulations is to give U.S. firms an advantage, to make sure that U.S. companies are leaders in AI and that the U.S. gets to write the rules of how AI will play out.

這些法規的目的是為美國公司提供優勢，確保美國公司在人工智能領域處於領先地位，並由美國來制定人工智能的發展規則。

Today, China is the world's largest importer of chips.

如今，中國已成為世界上最大的芯片進口國。

They spend as much money each year importing chips as they spend importing oil.

他們每年進口芯片的花費與進口石油的花費相當。

There's nothing that China's more reliant on the outside world to purchase.

中國沒有什麼比對外採購更依賴的了。

Right now, the most advanced Chinese firm, SMIC, is about five years behind TSMC, which might not sound like a lot, but that's two and a half Moore's laws behind TSMC, which means that for the most cutting-edge applications, you really take a performance hit if you want to use a Chinese manufacturer versus a Taiwanese one.

目前，最先進的中國公司中芯國際（SMIC）比臺積電（TSMC）落後大約五年，這聽起來似乎不多，但這比臺積電落後兩個半摩爾定律，這意味著對於最尖端的應用，如果你想使用中國製造商而不是臺灣製造商，你確實會在性能上受到影響。

That's the U.S. goal, to throw sand in the gears of China's AI ecosystem and hope that the U.S. can race ahead as a result.

這就是美國的目標，在中國人工智能生態系統的齒輪上摻沙子，希望美國能是以領先。

The biggest change in the past couple of years has been the explosion of investment in AI.

過去幾年最大的變化是對人工智能的投資激增。

The release of ChatGPT in late 2022 encouraged all the big tech firms to spend tens of billions of dollars building vast AI infrastructure, which means data centers full of the most capable semiconductors.

2022 年末發佈的 ChatGPT 鼓勵所有大型科技公司斥資數百億美元建設龐大的人工智能基礎設施，這意味著數據中心將佈滿功能最強大的半導體。

One of the key trends in the history of AI is that more advanced systems require being trained on larger volumes of data.

人工智能發展史上的一個重要趨勢是，更先進的系統需要在更大的數據量上進行訓練。

If you want to train a system on more data, you need more computing power, which means better chips to train it.

如果要在更多數據上訓練系統，就需要更強的計算能力，這意味著需要更好的芯片來訓練它。

So today, companies like OpenAI or Anthropic are spending millions and millions and soon billions of dollars training their AI systems, and most of that budget goes to buying chips, buying ultra-advanced semiconductors from companies like NVIDIA.

是以，今天，像 OpenAI 或 Anthropic 這樣的公司正在花費數百萬、數百萬甚至很快就會花費數十億美元來訓練他們的人工智能系統，而這些預算的大部分都用於購買芯片，從英偉達（NVIDIA）這樣的公司購買超先進的半導體。

One of the key challenges of AI is going to be to drive down the cost of deploying AI systems.

人工智能的主要挑戰之一是降低部署人工智能系統的成本。

To make AI really widespread across the economy, we need the cost of using it to be so cheap, we don't even think about it.

要想讓人工智能在整個經濟領域真正得到普及，我們需要讓人工智能的使用成本變得如此低廉，以至於我們想都不用想。

It's sort of like Google search today.

這有點像現在的谷歌搜索。

No one thinks, what's the price of my Google search?

沒有人會想，我在谷歌搜索的價格是多少？

Because it's approximately zero.

因為它大約為零。

Google spends a bit of money on the data centers, but it's so low, you don't have to think about it.

谷歌在數據中心上花了一些錢，但費用很低，你根本不用考慮。

Today, AI is actually pretty expensive.

如今，人工智能實際上非常昂貴。

A single query to ChatGPT is an appreciable amount of money.

ChatGPT 的一次查詢就是一筆可觀的費用。

There are a lot of companies that are exploring how do you do deployment more efficiently.

很多公司都在探索如何更有效地進行部署。

NVIDIA's chips, which are at the center of the AI ecosystem right now, are pretty general purpose in their capabilities.

英偉達™（NVIDIA®）芯片是目前人工智能生態系統的核心，其功能非常通用。

They can train many different types of models and are useful both for training and also for deployment.

它們可以訓練多種不同類型的模型，對訓練和部署都很有用。

But if you design a chip for a specific type of model or a specific type of deployment, you can make it perfectly optimized for that use case.

但是，如果你為特定類型的機型或特定類型的部署設計芯片，你就能使其針對該用例進行完美優化。

And so a lot of startups right now are looking at individual workloads or individual deployment opportunities and saying, we're going to design a chip that's perfectly tweaked for that use case.

是以，現在很多初創公司都在關注單個工作負載或單個部署機會，並表示，我們要設計一款針對該用例進行完美調整的芯片。

This is startups tackling this industry, but it's also big tech companies, Facebook, Microsoft, Google, they're all designing their own in-house chips as well, because they know the specific workloads that are inside their data centers.

這是初創企業在應對這個行業，但也有大型科技公司，如 Facebook、微軟、谷歌，它們也都在設計自己的內部芯片，因為它們知道數據中心內的特定工作負載。

And they've realized if they design chips specifically around those workloads, they can operate more efficiently in many cases than a general purpose AI chip like NVIDIA's can do, which I think is going to be really important in making AI cheap enough and therefore prolific enough to make a major impact on the economy.