Have we reached the limit of computer power? - Sajan Saini and George Zaidan

譯者: Lilian Chiu 審譯者: 麗玲 辛

In the Netherlands, there’s an ambitious company

在荷蘭，有一家野心勃勃的公司，

that builds one of the most advanced and expensive tools in the world:

打造了世界上最先進 且最昂貴的工具之一：

a single unit costs hundreds of millions of dollars.

單單一台的成本就高達數億美金。

And when companies buy one,

如果買了一台，

they also need 250 engineers to install the 165-ton device

還需要兩百五十名工程師

來安裝這個重達 一百六十五公噸的裝置，

in a process that typically takes half a year.

安裝過程通常需要半年。

But despite this steep cost in time and money,

但，儘管時間和金錢成本如此高昂，

many microchip makers desperately want one of these machines.

許多晶片製造商仍然 非常希望擁有一台這種機器。

The hundred-million-dollar question is: why?

價值上億美金的 重要問題是：為什麼？

The answer has to do with something called Moore’s Law.

答案與所謂的摩爾定律有關。

First coined by Intel co-founder Gordon Moore,

創造這個詞的人是英特爾的 共同創辦人高登‧摩爾，

this law states that every 1 to 2 years

摩爾定律指出，每一到兩年，

the number of transistors that can fit on a given size computer chip will double.

同樣尺寸的電腦晶片， 能容納的電晶體數量會加倍。

And by extension, the rough number of calculations that chip can do per second

進而，該晶片每秒 可執行的運算次數大致上

will also double.

也會加倍。

Now, this law isn't a physical law like gravity.

摩爾定律不是像重力 那樣子的物理定律。

It’s just a trend Moore observed during the early 1960s.

它只是摩爾在 1960 年代 初期觀察到的趨勢。

But chipmakers turned that trend into a goal, and in turn,

但晶片製造商將這種趨勢當成目標，

因此，消費者也開始期待運算技術

consumers learn to expect computing progress to continue

也會像這樣以指數級的 速度不斷進步。

at this exponentially fast pace.

And the amazing thing is, for six decades, it pretty much has.

驚人的是，六十年來， 事實也差不多是這樣。

Thanks to Moore’s Law, chips have gotten smaller, faster,

多虧摩爾定律，晶片變得更小、

more efficient, and cheaper.

更快、更有效率，且更便宜。

But today, there are four key problems that trip up this trend,

但現今，有四個重大問題 在阻礙這項趨勢，

potentially ending Moore’s Law

有可能會打破摩爾定律，

and fundamentally changing how we make progress in computing.

並徹底改變我們在運算方面 取得進展的方式。

The first is transistor size.

第一個問題是電晶體的尺寸。

Transistors are basically on/off switches,

電晶體基本上是個開／關裝置，

and these building blocks of digital computing have been shrinking

是數位運算的基礎元件，

since the 1960s.

從 1960 年代起， 它們就不斷在縮小。

But recently they’ve gotten so small,

但最近，它們已經小到

quantum physics has begun to interfere with their functions.

功能會受到量子物理干擾的程度。

When a transistor’s switch, or gate, is less than 20 nanometers,

當電晶體的開關或閘極 小於二十奈米時，

electrons will tunnel along it continuously,

電子會持續進行穿隧效應傳輸，

turning a crisp on/off switch into a hazy dimmer.

將本來清楚分明的開／關 變成模糊的調光器。

The second problem is heat.

第二個問題是熱。

As chipmakers make components smaller and more complex,

隨著晶片製造商把元件 做得更小、更複雜，

the copper lines that run between them need to be thinner and longer.

它們之間的銅線就得要更細、更長。

This increases their electrical resistance and generates high heat

這會增加它們的電阻，產生高熱，

that impairs chip performance and can’t be easily dissipated.

造成晶片性能降低， 且這高熱還很不容易消散。

Today’s chips can already run hot enough to cook an egg,

現今的晶片溫度 已經高到可以煎蛋了，

and temperatures are only predicted to increase without new innovations.

若沒有創新，預期 溫度只會繼續增加。

While both these issues represent limits in the fundamental physics of chipmaking,

雖然這兩個問題都代表 晶片製造中的基本物理學極限，

researchers haven't stopped trying to solve them.

但研究人員還在嘗試解決它們。

Unfortunately, their solutions often exacerbate the third major problem:

不幸的是，他們的解決方案 通常會加劇第三個重大問題：

chipmaking’s environmental impact.

晶片製造對環境的影響。

For example, swapping copper lines for ruthenium

比如，將銅線換成銠

could help pack transistors more tightly and keep chips smaller—

可以讓電晶體的排列更緊密， 維持晶片的小尺寸——

but that metal is far scarcer than copper

但這種金屬比銅更稀少，

and would require new mining infrastructure.

會需要新的採礦基礎設施。

Similarly, the technology currently used to make today’s smallest transistors

同樣的，要製造現今最小的 電晶體，目前所用的技術

requires huge amounts of energy and chemicals

會需要大量的能量

和化學物質：全氟烷基和 多氟烷基物質，

called perfluoroalkyl and polyfluoroalkyl substances,

which can take thousands of years to break down in the environment.

這些物質要數千年的時間， 才能在環境中分解。

Managing these first three problems contributes to the final issue: cost.

在解決前三個問題時， 卻會造成最後一個問題：成本。

To keep achieving Moore’s Law,

若要繼續實現摩爾定律，

chipmakers have to make individual chip components smaller.

晶片製造商必須把 每個晶片組件都做得更小。

And this is where that costly $400 million machine comes in.

此時就該四億美金的 昂貴機器登場了。

This marvel of chip-making science shoots a stream of tin droplets

這項晶片製造的科學奇蹟 會將一束錫滴射入真空室，

into a vacuum chamber before blasting them with a high-energy laser

然後用高能雷射 對它們進行強力照射，

that vaporizes the tin to create plasma.

把錫蒸發成為電漿。

In turn, the plasma emits a 13.5 nanometer wavelength of ultraviolet light

接著，電漿會發出 十三點五奈米波長的紫外光，

that can be used to produce incredibly small transistors.

可以用來製造極小的電晶體。

This remarkable feat of engineering has helped chipmakers

這項卓越的工程壯舉一直在協助 晶片製造商實現摩爾定律。

keep up with Moore’s Law.

But as chips keep getting denser,

但隨著晶片密度越來越高，

intricate manufacturing plants keep getting more expensive.

精密的製造工廠成本不斷攀升。

This trend has been so consistent, it’s actually earned the nickname

這種趨勢始終如一， 因此得到一個暱稱：

Moore’s Second Law.

摩爾第二定律。

Obviously, all these trajectories are unsustainable.

很顯然，這些發展路徑 不可能永遠持續下去。

Manufacturing plants can’t keep increasing in price,

製造工廠不可能不斷提升價格，

our ecosystems can’t endure endless mining and pollution,

我們的生態系統無法承受 無止盡的採礦和污染，

and the laws of physics are unlikely to change anytime soon.

物理定律短期內 也不太可能有所改變。

Fortunately, Moore's Law is flexible,

幸運的是，摩爾定律是有彈性的，

and there’s no reason we can’t introduce new goals

我們當然也可以引入新的目標，

to keep making computing progress responsibly.

用負責任的方式繼續推動 運算技術的進步。

Perhaps we could introduce a new Sustainability Law?

也許我們可以採用新的永續定律？

Smaller transistors already use less material and produce less e-waste,

較小的電晶體

使用的材料已經比較少了， 產生的電子廢物也比較少，

and advancements in electronic-photonic integration

而且電子與光子整合的進展

are allowing chips to use less energy and generate less heat.

讓晶片需要使用的能源更少， 產生出的熱也更少。

So perhaps chips should be made twice as sustainable every several years?

所以，也許每幾年， 晶片的永續程度就應該加倍？

Whatever the answer is,

不論答案是什麼，

we make the laws, so the future is up to us.

定律是我們制訂的， 所以未來也取決於我們。