The self-assembling computer chips of the future | Karl Skjonnemand

譯者: Lilian Chiu 審譯者: 易帆 余

Computers used to be as big as a room.

以前，電腦的大小跟 一個房間一樣大。

But now they fit in your pocket,

但現在已經小到可以放到 口袋裡、戴在手腕上，

on your wrist

and can even be implanted inside of your body.

甚至可以植入你的身體中。

How cool is that?

多酷啊？

And this has been enabled by the miniaturization of transistors,

這之所以能夠實現， 是因為電晶體的微型化，

which are the tiny switches in the circuits

電晶體是電路中的小型開關，

at the heart of our computers.

位在電腦的心臟部位。

And it's been achieved through decades of development

微型化能成功，也是經過數十年

and breakthroughs in science and engineering

科學和工程的的發展及突破，

and of billions of dollars of investment.

還有數十億美元的投資。

But it's given us vast amounts of computing,

但，它給了我們非常大量的計算、

huge amounts of memory

非常大量的記憶體，

and the digital revolution that we all experience and enjoy today.

以及我們現今大家都體驗到 且很享受的數位革命。

But the bad news is,

但，壞消息是，

we're about to hit a digital roadblock,

我們很快就要碰到數位路障了，

as the rate of miniaturization of transistors is slowing down.

因為電晶體微型化的 速度正在減緩。

And this is happening at exactly the same time

這個狀況發生的同時，

as our innovation in software is continuing relentlessly

因人工智慧以及大數據，

with artificial intelligence and big data.

我們的軟體還在持續不斷創新。

And our devices regularly perform facial recognition or augment our reality

且我們的裝置經常要執行 臉孔辨識或是虛擬實境，

or even drive cars down our treacherous, chaotic roads.

或甚至要在我們變化莫測 又混亂的道路上自動開車。

It's amazing.

這很不可思議。

But if we don't keep up with the appetite of our software,

但，如果我們趕不上 我們軟體的胃口，

we could reach a point in the development of our technology

我們的科技發展就有可能 會達到一個點，

where the things that we could do with software could, in fact, be limited

在這個點，我們用軟體能夠做的事

其實會受限於我們的硬體。

by our hardware.

We've all experienced the frustration of an old smartphone or tablet

我們都遇過這樣的挫折： 老式手機或平板電腦

grinding slowly to a halt over time

跑得又慢又辛苦，最後停下來，

under the ever-increasing weight of software updates and new features.

因為裝在上面的軟體更新 和新功能帶來的負擔越來越大。

And it worked just fine when we bought it not so long ago.

但不久前我們剛買來的時候 用起來還挺好的。

But the hungry software engineers have eaten up all the hardware capacity

但飢渴的軟體工程師

隨時間吃光了全部的硬體能力。

over time.

The semiconductor industry is very well aware of this

半導體產業非常清楚這個狀況，

and is working on all sorts of creative solutions,

且在努力投入各種創意解決方案，

such as going beyond transistors to quantum computing

比如超越電晶體，採用量子計算，

or even working with transistors in alternative architectures

或甚至在替代架構當中 使用電晶體，

such as neural networks

比如類神經網路，

to make more robust and efficient circuits.

以製造出更穩健且有效率的電路。

But these approaches will take quite some time,

但這些方法都要花相當的時間，

and we're really looking for a much more immediate solution to this problem.

針對這個問題，我們真的期望 能有更立即的解決方案。

The reason why the rate of miniaturization of transistors is slowing down

電晶體微型化的速度 之所以慢下來的原因

is due to the ever-increasing complexity of the manufacturing process.

是因為製程的複雜度不斷增加。

The transistor used to be a big, bulky device,

電晶體以前是大型笨重的裝置，

until the invent of the integrated circuit

直到以純晶體矽晶圓為基礎的

based on pure crystalline silicon wafers.

積體電路被發明出來。

And after 50 years of continuous development,

持續發展了五十年後，

we can now achieve transistor features dimensions

現在我們可以把電晶體尺寸

縮小到只有十奈米。

down to 10 nanometers.

You can fit more than a billion transistors

你可以把超過十億個電晶體

in a single square millimeter of silicon.

放入一平方毫米的矽當中。

And to put this into perspective:

更清楚來說，

a human hair is 100 microns across.

一根人類頭髮的寬度是一百微米。

A red blood cell, which is essentially invisible,

一個紅血球細胞， 基本上是看不見的，

is eight microns across,

寬度是八微米，

and you can place 12 across the width of a human hair.

所以，一根人類頭髮的寬度 約可放十二個紅血球細胞。

But a transistor, in comparison, is much smaller,

但，相對之下，電晶體更小，

at a tiny fraction of a micron across.

寬度只有一微米的一小部分。

You could place more than 260 transistors

大約兩百六十個電晶體排在一起

across a single red blood cell

才等同一個紅血球細胞的寬度，

or more than 3,000 across the width of a human hair.

或者，三千個電晶體排在一起， 才等同一根人類頭髮的寬度。

It really is incredible nanotechnology in your pocket right now.

現在就在你口袋裡的奈米科技 真的很不可思議。

And besides the obvious benefit

明顯的益處是能夠

of being able to place more, smaller transistors on a chip,

在晶片上放更多較小的電晶體，

smaller transistors are faster switches,

此外，較小的電晶體 也是較快的開關，

and smaller transistors are also more efficient switches.

且較小的電晶體也是 比較有效率的開關。

So this combination has given us

所以，這種組合讓我們可以

lower cost, higher performance and higher efficiency electronics

取得成本較低、性能較佳、 效率較高的電子產品，

that we all enjoy today.

讓我們現今可以享用。

To manufacture these integrated circuits,

要製造積體電路，

the transistors are built up layer by layer,

電晶體要一層一層打造

on a pure crystalline silicon wafer.

在純晶體矽晶圓上。

And in an oversimplified sense,

用極度簡化的方式來表示，

every tiny feature of the circuit is projected

電路的每一項小特徵都會被投影

onto the surface of the silicon wafer

到矽晶圓的表面上，

and recorded in a light-sensitive material

記錄在光敏感的材料中，

and then etched through the light-sensitive material

接著透過光敏感的材料進行蝕刻，

to leave the pattern in the underlying layers.

在下方的各層留下圖案。

And this process has been dramatically improved over the years

這些年來，這個流程 已經被大大地改善，

to give the electronics performance we have today.

讓我們現今使用的電子產品 能有這樣的效能。

But as the transistor features get smaller and smaller,

但，隨著電晶體的特徵 變得越來越小，

we're really approaching the physical limitations

我們已經越來越接近

這項製造技術的實體極限。

of this manufacturing technique.

The latest systems for doing this patterning

做這種曝影的最新系統

have become so complex

已經複雜到

that they reportedly cost more than 100 million dollars each.

據稱每台機器的成本要超過一億美元。

And semiconductor factories contain dozens of these machines.

半導體工廠有數十台這類機器。

So people are seriously questioning: Is this approach long-term viable?

所以，大家會質疑： 長期來看，這種方式可行嗎？

But we believe we can do this chip manufacturing

但，我們相信我們可以用完全不同

in a totally different and much more cost-effective way

且更有成本效益的方式來製造晶片，

using molecular engineering and mimicking nature

將分子工程以及模仿自然的方式

down at the nanoscale dimensions of our transistors.

運用到我們奈米尺度的電晶體上。

As I said, the conventional manufacturing takes every tiny feature of the circuit

如我前面說過的，傳統製造方式 是把電路的微小特徵

and projects it onto the silicon.

投射到矽上面。

But if you look at the structure of an integrated circuit,

但如果你去看積體電路的結構，

the transistor arrays,

電晶體陣列，

many of the features are repeated millions of times.

許多特徵其實被重覆了數百萬次。

It's a highly periodic structure.

它是種高度週期性的結構。

So we want to take advantage of this periodicity

所以，我們想要把這種週期性

in our alternative manufacturing technique.

應用到我們的替代製造技術。

We want to use self-assembling materials

我們想要用自組裝的材料，

to naturally form the periodic structures

來自然形成我們的電晶體

that we need for our transistors.

所需要的週期性結構。

We do this with the materials,

我們用適當的材料，

then the materials do the hard work of the fine patterning,

由材料來做難做的精緻圖形，

rather than pushing the projection technology to its limits and beyond.

而不是把投影技術 推到極限或極限之外。

Self-assembly is seen in nature in many different places,

在大自然的許多地方 都可以看到自我組裝的例子，

from lipid membranes to cell structures,

從脂質膜到細胞結構，

so we do know it can be a robust solution.

因此，我們知道它可以是個 穩健的解決方案。

If it's good enough for nature, it should be good enough for us.

如果它對大自然來說夠好了， 那對我們來說應該也夠好了。

So we want to take this naturally occurring, robust self-assembly

所以我們想要把這種大自然本有的 穩健的自我組裝方法

and use it for the manufacturing of our semiconductor technology.

用在製造半導體的技術上。

One type of self-assemble material --

其中一種自組裝材料——

it's called a block co-polymer --

叫做嵌段共聚物——

consists of two polymer chains just a few tens of nanometers in length.

含有兩個聚合物鏈， 長度只有幾十奈米。

But these chains hate each other.

但這些鏈痛恨彼此。

They repel each other,

它們會互相排斥，

very much like oil and water or my teenage son and daughter.

很像油和水，或是 我十幾歲的兒子和女兒。

(Laughter)

（笑聲）

But we cruelly bond them together,

我們用蠻力將它們結合在一起，

creating an inbuilt frustration in the system,

由於它們彼此互斥，

as they try to separate from each other.

所以就就形成了內建的阻撓系統。

And in the bulk material, there are billions of these,

有大批這樣的材料，有數十億種，

and the similar components try to stick together,

類似的材料試圖黏合在一起，

and the opposing components try to separate from each other

而與此同時，對立的材料 則試圖與彼此分開。

at the same time.

And this has a built-in frustration, a tension in the system.

這系統內建著阻撓與拉力。

So it moves around, it squirms until a shape is formed.

它會到處移動、蠕動，直到成形。

And the natural self-assembled shape that is formed is nanoscale,

自然自組的形狀小到奈米級，

it's regular, it's periodic, and it's long range,

它有規律，有週期性，範圍很長，

which is exactly what we need for our transistor arrays.

正如電晶體陣列所需。

So we can use molecular engineering

這樣我們就可以使用分子工程

to design different shapes of different sizes

來設計不同大小的不同形狀，

and of different periodicities.

以及不同的週期。

So for example, if we take a symmetrical molecule,

比如，以一個對稱的分子為例，

where the two polymer chains are similar length,

在這個分子中， 兩個聚合物鏈的長度相近，

the natural self-assembled structure that is formed

形成的自然自組結構

is a long, meandering line,

是一條很長且蜿蜒的線，

very much like a fingerprint.

非常像是指紋。

And the width of the fingerprint lines

而指紋線的寬度

and the distance between them

和它們之間的距離

is determined by the lengths of our polymer chains

是根據我們聚合物鏈的 長度來決定的，

but also the level of built-in frustration in the system.

此外系統內建的阻撓程度 也是一個決定因子。

And we can even create more elaborate structures

如果我們能使用不對稱的分子，

我們甚至可以創造出 更精緻的結構，

if we use unsymmetrical molecules,

where one polymer chain is significantly shorter than the other.

不對稱的意思就是 兩條聚合物鏈的長度明顯不同。

And the self-assembled structure that forms in this case

在這個情況下形成的自組裝結構，

is with the shorter chains forming a tight ball in the middle,

比較短的鏈會在中心 形成一個緊實的球，

and it's surrounded by the longer, opposing polymer chains,

它的周圍則是較長、 對立的聚合物鏈，

forming a natural cylinder.

形成一個自然的圓柱。

And the size of this cylinder

這個圓柱的大小

and the distance between the cylinders, the periodicity,

以及圓柱間的距離，即週期性，

is again determined by how long we make the polymer chains

同樣也是取決於我們 製造的聚合物鏈的長度，

and the level of built-in frustration.

以及內建的阻撓程度。

So in other words, we're using molecular engineering

換言之，我們用分子工程

to self-assemble nanoscale structures

來自組奈米尺度的結構，

that can be lines or cylinders the size and periodicity of our design.

可以根據我們的設計來形成線條、 圓柱大小和週期不同的結構。

We're using chemistry, chemical engineering,

我們利用化學、化學工程，

to manufacture the nanoscale features that we need for our transistors.

將我們需要的奈米特性 製作在電晶體上。

But the ability to self-assemble these structures

但，自主組裝這些結構的能力

only takes us half of the way,

只能帶我們走到半路，

because we still need to position these structures

因為我們仍然需要 將這些結構放置在適當的位置，

where we want the transistors in the integrated circuit.

而這些位置，就是我們希望 電晶體在積體電路中擺放的地方。

But we can do this relatively easily

但我們能相對輕易地做到，

using wide guide structures that pin down the self-assembled structures,

只要運用大範圍的指引結構，

將自組裝結構引導到

anchoring them in place

我們指定的固定位置，

and forcing the rest of the self-assembled structures

迫使其餘的自組結構平行排列，

to lie parallel,

aligned with our guide structure.

如此就能依照我們的建構方式 完成結構的組建。

For example, if we want to make a fine, 40-nanometer line,

比如我們想做一條 四十奈米長的細線，

which is very difficult to manufacture with conventional projection technology,

很難用傳統的投影技術來製造，

we can manufacture a 120-nanometer guide structure

但我們可以製造一個 一百二十奈米的結構引導通道，

with normal projection technology,

用一般的投影技術就辦得到，

and this structure will align three of the 40-nanometer lines in between.

這個通道結構中會有

三條四十奈米互相對齊的線。

So the materials are doing the most difficult fine patterning.

如此，材料才能完成 最困難的精緻曝影。

And we call this whole approach "directed self-assembly."

我們把這整個方法叫做 「引導式自組裝」。

The challenge with directed self-assembly

引導式自組裝的挑戰在於

is that the whole system needs to align almost perfectly,

整個系統需要近乎完美地 符合我們要的排列方式，

because any tiny defect in the structure could cause a transistor failure.

因為結構中若有任何微小的瑕疵，

都可能會造成電晶體故障。

And because there are billions of transistors in our circuit,

因為我們的電路上 有數十億個電晶體，

we need an almost molecularly perfect system.

我們需要一個接近 分子等級的完美系統。

But we're going to extraordinary measures

但我們需要用到非常精準的量測工具 才能達成這個目標，

to achieve this,

from the cleanliness of our chemistry

從化學的清潔，

to the careful processing of these materials

到半導體工廠小心處理這些材料，

in the semiconductor factory

to remove even the smallest nanoscopic defects.

到移除最小的奈米尺度瑕疵。

So directed self-assembly is an exciting new disruptive technology,

所以，引導式自組裝是種 讓人興奮的顛覆性新技術，

but it is still in the development stage.

但它還在開發階段。

But we're growing in confidence that we could, in fact, introduce it

但我們越來越有信心可以真的

把它引入到半導體產業，

to the semiconductor industry

as a revolutionary new manufacturing process

做為一種革命性的新製程，

in just the next few years.

且在接下來幾年就可以做到。

And if we can do this, if we're successful,

如果我們能做到，如果我們成功，

we'll be able to continue

我們將能夠把電晶體的 成本效益繼續微型化 ，

with the cost-effective miniaturization of transistors,

continue with the spectacular expansion of computing

繼續將計算能力大大擴展，

and the digital revolution.

並帶來數位革命。

And what's more, this could even be the dawn of a new era

不只如此，這甚至可能是

分子製造新紀元的黎明。

of molecular manufacturing.

How cool is that?

這多酷啊？

Thank you.

謝謝。

(Applause)

（掌聲）