Is quantum supremacy (the moment traditional computers become obsolete) fiction?
In quantum computing, many consider quantum transcendence to be a true breakthrough, but is it really a mirage?
“Quantum transcendence”-Even this phrase gives the impression of self-awareness, almost like a terminator. It’s a moment when everything changes, traditional computers become obsolete, and quantum computers take over.
The reality is more subtle. Quantum supremacy represents the latest in some milestones in quantum computing. This led us from the first idea of the 1980s that quantum mechanics could be used to build very powerful types of computers. I realized in the 1990s that I could correct quantum errors. Demonstration of qubits on many platforms in the 2000s. And for the construction of quantum integrated circuits in the 2010s.
But what quantum computing insiders know is that the “basic” quantum supremacy that can be achieved in less than 100 qubits does not necessarily mean that it solves a useful problem. about it.
What is hegemony?
Quantum supremacy is what seems to be virtually impossible on a traditional computer, either because it can take thousands of years to perform a calculation, or because it lacks all the available memory. Achieved when the computer can run. The catch is that this task doesn’t have to be useful at all. Indeed, the first demonstration of hegemony is completely useless in itself.
The more relevant questions are: Once the advantage is achieved, what is the next path to useful quantum computing and how do we get there?
There is optimism that quantum software identifies early-stage quantum processors that have not been error-corrected and are used to provide useful solutions. Quantum software start-ups are struggling to lead the development of “quantum apps” to unlock the potential of computing.
Despite this optimism, it’s still unclear if there are any really useful issues that these early-stage machines can solve. Modeling materials and chemicals can be exciting, for example using a small quantum processor that runs in conjunction with a traditional computer, but even with these, a huge number run in parallel. Quantum coprocessor is expected to be needed.
These “Noisy Medium Quantum” (NISQ) processors are limited because they are designed to simply tolerate and avoid errors rather than fix them. We are far more confident in the capabilities of quantum computers with built-in error correction, but they require more than 1000 physical qubits for each pure error-corrected “logical qubit”. It can take millions of qubits to get enough error-correcting qubits to run a general-purpose quantum algorithm.
NISQ processors play an important role in exploring and developing quantum apps and understanding the capabilities of the technology, but look beyond this era to quantum processor technology that can scale to over a million cubics and integrate well. It is also essential to point it. It uses a traditional computer and has built-in density and error immunity.
Scaling with silicon
The design of a million qubit quantum chip seems to be a major leap from the 50-100 qubit processor currently under development. But look at the silicon transistor for comparison. Today, one chip has billions instead of millions. By 2025, it is predicted that there will be more silicon transistors on Earth than human cells.
Building qubits from silicon transistors, or devices very similar to them, offers great scaling potential by benefiting from mature silicon manufacturing processes and equipment.
The benefits of silicon qubits go beyond taking advantage of the mature $ 1 trillion manufacturing industry. Silicon also provides some with the lowest inherent error rates, increasing the capacity of the physical qubits so that more can be devoted to useful calculations instead of fixing the errors.
Silicon cubits already show early star power in this regard. Their quantum states have a lifetime of more than a second, which is longer than any other cubit in solid-state devices. Much of this results in chemistry. The main isotope of silicon is non-magnetic, allowing a quiet environment for silicon cubics, and this refined silicon 28 isotope “large” 300mm wafer is already manufactured in an industrial grade facility. .. Ready to process using standard CMOS manufacturing methods.
Another important advantage of silicon qubits is their high density. This could accommodate millions of qubits in 1 cm.2The ability to integrate the chip, and traditional coprocessors and control logic directly into the same chip as the quantum core.
Silicon quantum chips still need to operate at ultra-low temperatures near absolute zero, and like today’s high-performance computing facilities, quantum computers are probably housed in data centers and accessed via standard cloud interfaces or APIs. is needed.
I’m looking forward to
Two eras are envisioned for silicon-based quantum processors:
- Multi-core NISQ era, Many NISQ processors (for example, hundreds of qubits) are tiled across the same chip. Even without complete quantum error correction, it can support machine learning and approximate optimization algorithms by asking the same question to thousands of identical quantum processors.
- The Age of Million QubitsWhen scalable manufacturing allows the quantum processor to be extended to thousands or millions of qubits. This will enable universal fault tolerant (UFT) quantum computing, which is expected to dramatically improve the ability to calculate solutions to problems such as material modeling for drug discovery and quantum chemistry.
The extraordinary progress we have witnessed in the development of quantum computing is the result of close collaboration between academic groups, microelectronics centers, national laboratories, and industry around the world.
There are many technical challenges on the road ahead. It’s a huge, complex and fascinating puzzle. The outlook for quantum computing has been unrecognized for five years and we are accelerating rapidly. It will be a pretty vehicle.
True quantum transcendence and competition for a million qubit chips
Source link True quantum transcendence and competition for a million qubit chips