The digitalization trend in almost all verticals — industrial, manufacturing, transportation, telecommunications, logistics, medical, research — has created an explosion of both structured and unstructured data. At the same time, advances in technology such as AI machine learning (ML), the cloud and now the metaverse are happening at an exponential rate. These advancements will soon enough require even more and more processing power than the latest supercomputers can provide. However, there are certain problems that even supercomputers aren’t very good at solving because they don’t have the working memory to hold the myriad combinations of real-world problems, and they have to analyze each combination one after another, which can take a long time.[1]
In comes quantum computing, a new computing paradigm that harnesses the power of quantum mechanics to deliver the ultimate in parallel computing.[2] Quantum computing has the potential to help industries, research institutions, and the society solve problems that currently overwhelm today’s classical computers. That said, to speed up the quantum leap, wafer testing solutions will also need to evolve to keep pace. Traditional methods are no longer cost and time effective to cater for a shorter time-to-market, which is key for businesses in today’s fast-changing environment.
To start off, quantum computing is based on a discovery that atoms do not follow traditional rules of physics.
Classical computers work with bits—consisting of ones and zeroes—but a quantum computer uses quantum bits or qubits, which, on top of the ones and zeroes, also has a superposition where it can be a one and a zero at the same time.
Because of this, operations on qubits can amount to a large number of computations in parallel, a characteristic called entanglement. If one qubit simultaneously represents two states, two qubits represent four states when coupled together. They can no longer be treated independently — they now form an ‘entangled’ super state.
As more qubits link together, the number of states exponentially increase, leading to a computer with astronomically large computing power. In theory, some specific problems are said to be solvable in much less time on a quantum computer than using the best-known algorithms for a classical computer.[3]
There are no commercial quantum computers yet, but development has advanced steadily over the past decade, with multiple companies now offering quantum applications as a service via cloud platforms such as Amazon Web Services, Google Cloud, and Microsoft Azure.[4]
According to market analyst International Data Corp. (IDC), customer spend for quantum computing is projected to grow from $412 million in 2020 to $8.6 billion in 2027—representing a six-year compound annual growth rate (CAGR) of 50.9% over the 2021-2027 forecast period.
The forecast includes core quantum computing as a service as well as enabling and adjacent quantum computing as a service.
Mainly driving this strong growth are major breakthroughs in quantum computing technology, a maturing quantum computing as a service infrastructure and platform market, and the growth of performance intensive computing workloads suitable for quantum technology.
IDC also expects investments in the quantum computing market—made by public and privately funded institutions, government spending worldwide, internal allocation (R&D spend) from technology and services vendors, and external funding from venture capitalists and private equity firms—to grow at a CAGR of 11.3% and reach nearly $16.4 billion by the end of 2027.
Key to the successful development of quantum computers is the quantum processor.
However, the inherent characteristics of qubits are making the development and manufacturing of such chips extremely complicated. For one, quantum states are fragile. They can collapse into a classical state if disturbed by noise or measurement.
Another issue is the no-cloning theorem, which means it is not possible to copy the state of one qubit onto another without altering the state of the first one, thereby depriving developers of the classical error-correction tool: copying.[5]
Moreover, the implementation of semiconductors in developing quantum chips has been challenging in its own right because some materials can exhibit many quantum degrees of freedom that may cause qubits to interact with each and decohere quickly.[6]
Of course, researchers and designers have already developed quantum computing semiconductor chips. But today’s quantum systems only have tens or hundreds of entangled qubits, limiting them from solving real-world problems.
Intel Corp. has a 49-qubit superconducting quantum test chip, while IBM Quantum’s 127-qubit Eagle quantum processor is already considered the ‘biggest’, and therefore, most powerful quantum chip available. For quantum computing to achieve practicality, commercial quantum systems need to scale to over a million qubits and overcome daunting challenges like qubit fragility and software programmability to make the leap from research to commercial viability in many applications.[7]
Bringing quantum computing from laboratory to production
Wafer probing technologies such as the Cryogenic Wafer Prober, a system designed and built as a collaboration between AEM and another Finnish based company, Bluefors can help speed up that leap.
Cryogenic wafer probing is important for device testing of several emerging technologies, such as cryogenic quantum computing and supra-conductive CMOS semiconductors, wherein temperatures near absolute zero are essential.
In order to reach these temperatures in a cost-effective and fast way, it is beneficial to cool the entire wafer inside a thermally and atmospherically isolated vacuum chamber.
Advanced 300mm wafer probe as the Cryogenic Wafer Prober system can operate under 4K and even below 2K—the necessary environment for testing quantum chips.
Traditionally, quantum chips are tested one by one. First, they are diced, cooled down to cryogenic temperatures, tested, and after this process, heated up again. And then on to the next die, until all dies from the wafer have been tested. The Cryogenic Wafer Prober is a system, where AEM provided the probing platform, while Bluefors provided the cryogenic system needed for the cooling.
What’s unique with solutions like the Cryogenic Wafer Prober system is that the whole wafer (150mm–300mm) can be probed and changed through the load lock in a cryogenic temperature, without heating the system to room temperature, thereby providing researchers much faster way of testing.
Specifically, Cryogenic Wafer Prober allows up to 100 times faster throughput in sample characterization of cryogenic quantum devices because all good dies in the wafer will already be characterized, dramatically speeding up the development of these devices.
On top of this, the Cryogenic Wafer Prober system enables users to go backwards in their process to be able to make adjustments and ensure higher yields in their next wafers.
Together with AEM’s low-temp probe cards, Cryogenic Wafer Prober’s unique active alignment system allows users to locate and contact devices automatically anywhere on the wafer.
Its modern and intuitive user interface provides direct control and full overview.
Meanwhile, the four servo-controlled axes of the probe station can be switched off during measurements to minimize electronic noise, and the load lock system allows fast wafer change in cryogenic temperatures.
When the customer has a full wafer and they don’t really know which devices are ready to go to market and which are not, this tool can characterize the whole wafer. For example, devices on the right side of the wafer, are better than the ones on the left side, or there are better devices on the center. With this tool the information about the whole wafer is available in just one to two days.
This information can be used to go backwards in the process, as the customers are creating their wafers and designs.
The tool allows to see which parts of the wafer are good, so the previous steps in the process can be readjusted to make better wafers that have a bigger set of good, working dies. In a way, effectively reducing defects at the end of the line.
Quantum computing has been proven to work, but an actual working quantum computer is not yet available.
Every once in a while, the industry will hear proclamations of quantum supremacy, but then again, the world is still far from quantum practicality.
Still, quantum computing research continues to open up new findings that will help the industry edge closer to commercial viability. And solutions such as the Bluefors and AEM Cryogenic Wafer Prober can be one of the tools to get the industry there.
For more Embedded, subscribe to Embedded’s weekly email newsletter.
You must Sign in or Register to post a comment.
This site uses Akismet to reduce spam. Learn how your comment data is processed.