量子位黑马 Dark Horse of Qubits
近几年，量子计算绝对是一个热门的研究领域。但正当所有科技巨头都在争夺创造宇宙最快计算机机会的时候，一个鲜为人知的公司 —— PsiQuantum已经在他们的C轮融资中获得1.5亿美元，并宣称他们能够在仅仅 5 年内作出许多量子专家预计需要 7 - 10 年或更多的成果。
Quantum computing has been a popular research field in recent years. While all the tech giants are competing for the fastest computer in the universe, a little-known company – PsiQuantum, has gathered $150 million in their Series C funding, and claims they will be able to do things in 5 years, for what many quantum experts predicted will take 7-10 years or more.
那么，这匹潜在的黑马究竟拥有了什么独一无二的秘密呢？我们得从神秘的光粒子开始，光子。
So, what is it this potential dark horse has but no other does? Well, we will have to start with the mystical light particles – photons.
图为IBM量子计算机的内部
The interior of IBM's Quantum Computer
希望之‘子’
Particle of Hope
自19世纪初以来，科学家们就已经开始研究光和电子之间的微妙关系。然而，直到1900年，在对黑体辐射的研究中，马克斯·普朗克才提出，电磁波所携带的能量只能以能量包的形式释放，后来1905年，爱因斯坦将其命名为光子。
Ever since the early 19th century, scientist have been studying the delicate relationships between light and electrons. However, it was not until 1900, during the study of black body radiation, Max Planck suggested that the energy carried by electromagnetic waves could only be released in packets of energy, which in 1905, Einstein named as photons.
在他的论文中，爱因斯坦提出：频率为𝜔的光其实是一束光量子，每个光量子都具有𝐸= ħ𝜔能量。（他后来也因此在1921年被授予诺贝尔奖，不是𝐸=𝑚𝑐^2哦）。
In his paper, Einstein proposed that light of frequency 𝜔 is a stream of light quanta, each with energy 𝐸 = ħ𝜔. (He is later awarded the Nobel Prize in 1921 for this, not 𝐸 = 𝑚𝑐^2.)
但是，在所有这些发现之前，光，或者更准确地说，电磁辐射已经被证明是一种波，并且正如以上所述，它们有着不同的频率。那么，这是否意味着他们当年的判断是错误的呢?
However, way before all these discoveries, light, or more precisely, electromagnetic radiations are proven to be waves, and as say above, they have different frequencies. So, does this mean they got it wrong back then?
波和粒子本质上的区别在于，波不一定受限于空间中的某部分，所以也可以受到干涉，而粒子则不能。但双缝实验的结果表明，光不但受到了干涉，还以局域粒子的形式出现了。所以光既是粒子又是波吗?
Well, the difference between waves and particles is that waves are not necessarily localized in space, and so can undergo interference, whereas particles can’t. During the double slit experiment, results show that light was interfered while still appearing as localized particles. So light is both particle and wave?
事实没那么简单！光的这种诡异行为仍然在被努力的研究和理解中，目前最好的理解是，光既不是波，也不是粒子，而是一种既像波又像粒子的东西!
The truth may not be so simple! While better understanding of this behavior is still undergoing investigation, for now, it is best to understand light as neither waves nor particles, but something that behave a bit like both!
图为波（蓝）和波粒（红）通过双缝实验干扰后在探测屏上累计的不同表现：波在干扰后显现出了的相长干涉和相消干涉部分，而光子在受到干扰后不仅显现了与波相似的干涉模式，还依然拥有着粒子的属性
The wave (blue) and wave-particle (red) showing their patterns on the screen after interference from the double slit experiment over time: wave shows the constructive and destructive interference patterns, while the photons show both the similar interference patterns to the wave and their particle properties
是现实，还是梦？
Reality or a Dream?
我是不是让你对光子的含义变得更晕了？不要担心，对于我们这些业余的来说，我们应该只想知道这些发现如何能让我们的生活变得更好吧。
Have I just made the meaning of photons more confusing? Not to worry. For us amateurs, we probably just want to know how those discovers can make our lives better.
回到我们之前提到的‘科幻’公司，PsiQuantum正是利用光子是粒子这个理论来执行量子计算的人- 硅光子。在传统计算机中，1和0的产生来自磁位，然后进行计算，量子计算机使用了各种其他技术来制造被称为量子位的‘磁位’，而正是这些量子位的叠加和纠缠使得信息处理的速度变得更快。
Going back to our almost science fictional company, PsiQuantum are the exact people using the fact that photons are particles – silicon photonics, to perform quantum calculations. While classical computers use magnetic bits to create the ones and zeros for computation, quantum computers use a variety of other technologies to make quantum bits called qubits. It is the superposition and entanglement of qubits that allows information processing to be so much faster.
PsiQuantum的目标是建立一个拥有一百万个量子位的量子计算机，而在量子位世界里那基本是个天文数字。举例来说，当今最大最好的量子计算机的量子位数不到100个。谷歌最近历史性的成就：在几秒钟内完成一项一台普通计算机数千年才能完成的计算任务，只用了54量子位。
PsiQuantum's goal is to build a quantum computer with a million qubits, and in the qubit world, that is an astronomical number. For perspective, today's biggest and best quantum computers have less than 100 qubits. Google’s recent historic achievement of quantum supremacy performance on a difficult computational task in a few seconds only took a mere 54 qubits. That would have taken a classical computer thousands of years to complete.
图为传统磁位（左）和量子位（右）的对比。在量子计算中，信息的基本单位是一个量子位，它可以有效的同时是1和0和其他之间的数字。
The comparison between classical bit (left) and qunit (rightist ). In quantum computing, the basic unit of information is a qubit, which can effectively be a one and a zero or something inbetween simultaneously.
那么冷，值得吗
So Cold, but Worth It
超导量子位是目前最流行的量子计算技术之一，是谷歌、英特尔、IBM和Rigetti所有量子发展的基础。这些设备基本上是在芯片上制造的小线圈，从而在芯片上复制了经典计算机中的线圈。然后当温度降到几乎绝对零度时（-270 °c左右），量子效应就产生了。
Superconducting qubits is most popular quantum computing technology used at the moment and is the foundation to all those developments with Google, Intel, IBM, and Rigetti. The devices are basically small coils fabricated on chips that replicating those in the classical computers. The quantum effect comes in when the temperature is brought down to almost absolute zero (around -270 °c).
在这些条件下，线圈会变成超导体，也就说流动的电流基本没有电阻了。它们的顺时针和逆时针流动将代表1，或0，或1和0之间的所有可能性的重合。
The coils in those conditions become superconductors, basically means resistance free for the currents flowing. Their clockwise and anti-clockwise flow will represent a one or a zero or a superposition of everything between one and zero.
虽然目前的芯片制造技术能够制造这些设备，也最常用的一个量子技术，但这种方式仍然有一些许弱点。首先，系统会很快失去它的量子态，这意味着被计算的问题所需要的步数就受到了限制。
Although current technology is able to manufacture these devices with chip fabrication techniques and is most commonly used one, there are a still few weaknesses to this method. First of all, the system losses its quantum state very quickly, meaning that there is a limitation on the number of steps requires to solve the problem.
其次，这里的量子位元只能接触到最近的其他量子位，因此要连接到稍微远一点的量子位的话，过程中必须进行多次连接。也因为使用了这些“踏脚石”，计算时间将会大大增加，从而限制了计算的复杂性。
Secondary, the qubits are only able to connect to their nearest qubit, therefore several connections would have to be made to reach a slightly more distant qubit. By using these ‘steppingstones’, there will be an increase in time of calculation hence limiting the complexity of the questions.
图为transmon量子位的模拟图像和扫描电镜图像，一种旨在降低对电荷噪声的敏感性的超导电荷量子位
The Optical and SEM images of a transmon qubit, which is a type of superconducting charge qubit designed to have reduced sensitivity to charge noise
被禁锢的孩子们
The Trapped Ones
最新的不一定是最好的。囚禁离子量子位的技术从90年代初就开始了。霍尼韦尔（Honeywell）和IonQ都是这种更大规模更商业量子技术的忠实粉丝，原子钟也使用了这种囚禁离子技术。
The new one is not always the best. The technology of trapped ion qubits has begun since early 1990s. While Honeywell and IonQ are both loyal fans of what has become a more commercial large-scale quantum technology, atomic clocks also use trapped ion technology.
霍尼韦尔和IonQ在他们的量子位技术都使用了镱（一种稀土金属）的同位素，但使用其他带电原子粒子也会有相似的效果。将激光精确的射于镱原子的外层电子，并将它们移除后，原先的材料粒子会变成离子。然后用激光再作为镊子来移动离子，直到所有粒子都就位。随后当所有的粒子都就位时，振荡的电压场会被开启并用来保持整个系统的状态。量子位会被储存在这些稳定的电子态离子中，而量子信息可以通过离子的集体量子化运动来传递。
For Honeywell and IonQ, their qubits both use the isotopes of the rare-earth metal called ytterbium, though other charged atomic particles can also be used. Precise lasers are applied to the outer electrons of the ytterbium atom and removes them, which turns the materials into ions. Lasers are then used as tweezers to move the ions around until all particles are in position. Once there, oscillating voltages fields would be used to keep ‘trapped’. The qubits would be stored in these stable electronic state ions, while quantum information can be transferred via the collective quantized motion of the ions.
与超导量子位相比，这里的离子能够维持其量子态的时间要长得多。量子态维持的时间越长，系统能够执行的计算就越复杂。因此，就在几天前，霍尼韦尔宣称他们已经创造了量子计算性能的新纪录，H1系统模型成为了第一个实现512个基准的量子容量（量子容量是量子计算机的能力和错误率的一种度量）的商业系统。
Compared to superconducting qubits, ion here are able to maintain their quantum state for a lot longer. The longer the quantum state maintained, the more complexed computation the system is able to perform. So just a few days ago, Honeywell claims that they have set new record for quantum computing performance, with System Model H1 becoming the first commercial system to achieve a quantum volume (a metric measurement of the capabilities and error rates of a quantum computer) of 512 benchmarks.
图为一串14个被禁锢后彼此纠缠的离子组
A string of 14 trapped and entangled ions
变个方向怎么样
Try Another Way
比起线圈和离子，PsiQuantum更加魔幻的计划是极化单光粒子。通过垂直偏振的光子将代表1，水平偏振光子将代表0，而对角偏振光子将表示1和0的叠加。
Instead of coils and ions, PsiQuantum’s plan is to polarize single light particles. By polarizing photons vertically, they will be able to represent one, polarizing horizontally will make zeros, while diagonally polarized photons will represent a superposition of both one and zero.
通常在我们日常生活中的光会在所有垂直于运动方向的方向上振荡。当极化作用于横波时，它会阻止某些方向上振荡的特定波，使波只在一个或几个方向上振荡。
Normally light in our daily life oscillate in all directions perpendicular to the direction of motion. When polarization is applied to transverse waves, it stops specific waves in some orientations, making the wave only oscillating in one or a few orientations.
这项量子技术的秘密来始于2009年PsiQuantum创始人Jeremy O'Brien的一篇研究论文。他相信，凭借“高速传输和光子出色的低噪声特性”，光子必将在量子计算领域扮演核心角色。
The secret to this technology came from a research paper in 2009 by the founder of PsiQuantum, Jeremy O’Brien. He believes with the ‘high-speed transmission and outstanding low-noise properties of photons’ the are bound to take on this central role of quantum computing.
使用光子来做量子位元的一个显著优势是，它们能够在其量子状态中停留相当长的时间。Lyman-alpha blob (LAB)就是一个很好的例子，它是一个会发射出了Lyman-alpha发射线的高浓度巨型气体，距离我们有115亿光年。经历了这么长时间的旅行，从那里来的光子到达地球时仍然保持着它们最初的极化状态。
A significant advantage of using photons for qubits is that they are able to stay in their quantum state for incredibly long. A good example is the Lyman-alpha blob (LAB), which is a huge concentration of a gas ~11.5 billion light years away, emitting the Lyman-alpha emission line. Having travelled for so long, photons from there still arrive on Earth with their original polarized states.
左图为在115亿光年之外的Lyman-alpha blob的光学图像，其中的黄色代表发光氢气。右图为是一个根据红外线数据的艺术创作，是在Lyman-alpha blob附近可能看到景象，如果看得比较近，其中白色为气团中的星系。
The left shows the Lyman-alpha optical image 11.5 billion light-years away, with yellow parts as glowing hydrogen gas in the blob. On the right is an artist's impression of what it might look like if viewed relatively close, a galaxy in the blob is shown as white.
光子量子位的大小非常小。作为波长只有几微米的它，这既是优点也是缺点。小波长将允许更好的计算精度，但它也使其更难控制。想象一下，试图操纵一个病毒大小并每秒3亿米运动的物体，何其之难啊。
The photon qubits will be very small. Having a wavelength of a few micrometres, it can be both an advantage and a disadvantage. The small wavelength would allow for better precision in the calculations, but it also makes it harder to control. Just imagen trying to manipulate something the size of a virus traveling at 300 million meters per second.
与超导线圈和静止离子不同，光子总是在以宇宙最快的速度运动。要同时处理上百万个躁动的光子将是一个巨大的挑战，更不用说读取、控制和操纵它们了。
Unlike the superconducting coils and the stationary ions, photons will always be moving as the fastest thing in the universe. It will be a challenge trying to juggle a million of those at a time, let alone reading, controlling and manipulating them as well.
在量子位世界中，还有一个PsiQuantum与许多其他研发者需要克服的问题，那就是数据错误修正。根据PsiQuantum的说法，这个项目很大一部分精力会用来解决这一问题。因此，最终很可能他们的百万量子位中的一大部分将用于错误监测和纠正。在今天的量子位技术中，每一个计算量子位都需要数千个纠错量子位。
Another big issue for PsiQuantum as the rest of the qubit world is error correction. However, according to PsiQuantum, they will be putting a significant amount of effort one this. Therefore, it is likely that a large proportion of their million qubits will be devoted to error monitoring and correcting. For today’s qubit technology, thousands of error correcting qubits are required for every computation qubit.
所以，虽然这种新型硅光子学技术听起来确实很有前途，但是前方依然有许多要克服的困难，那我们就在5年后见分晓吧。不过，微软已经作为战略投资者支持他们了哟。
So, this new silicon photonics technology does sound promising, but with many challenges ahead, we will see how it actually turns out in 5 years. Bearing in mind, Microsoft has already backed them up as a strategic investor.
即使PsiQuantum最终只能
创造一个修正错误的千个量子位计算机，
他们也将创造一个改变世界的历史时刻。
它将帮助基础科学、医学和许多其他领域
取得数千项突破。
Even if PsiQuantum ended up with
a thousand error corrected qubits,
they will have created
a world changing historical moment.
Making thousands of breakthroughs in
fundamental science, medicine and many other fields.
图为在PsiQuantum的帕洛阿尔托(Palo Alto)实验室，正在进行研究的硅片
A PsiQuantum silicon wafer at the company’s lab in Palo Alto during research
图片来自 Forbs, ITIF, Bloomberg, Wikipedia官网以及以下文章
文中部分英文信息参考来自 Paul Smith-Goodson 在福布斯的 ‘Quantum Computing With Particles Of Light: A $215 Million Gamble‘ 文章
其余中英文内容为原创
Images from official website of Forbs, ITIF, Bloomberg, Wikipedia and the below articles
Parts are sited from ‘Quantum Computing With Particles Of Light: A $215 Million Gamble‘ from Paul Smith-Goodson
The rest of the Chinese and English content are original
