Google has announced that it has made a major breakthrough in quantum computing, possibly pushing the technology from conceptual to practical.
For the first time, the tech company says it has developed a state-of-the-art quantum computing chip called Willow that can solve, in less than five minutes, a calculation so complex, it would take about 10 septillion (10^25) supercomputers. ) to complete the year.
“The Willow chip is a major step in a journey that began 10 years ago,” Hartmut Neven, the Google executive who founded and led Quantum AI, the research team behind the breakthrough, said in a blog post on Monday, December 9.
Google also said it has discovered how to reduce errors in quantum computers by using more qubits to scale the technology — something that has eluded researchers in the domain for the past 30 years.
“We’ve always hypothesized that a quantum computer can do something that a classical computer can’t. That’s the main purpose of quantum computing, but it’s based on a theoretical construct. If Google’s claim is true, it’s a demonstration that that hypothesis is correct,” Debpriya Basu Roy, Assistant Professor, Department of Computer Science, Indian Institute of Technology (IIT) Kanpur said. indianexpress.com.
“We all know the potential of quantum computers, but making computers practical and solving some real-world problems is a significant step forward,” he added.
Let’s take a closer look at the essentials of quantum computing, what Google has achieved with its new Willow chip, and whether it stands to gain an edge in the ongoing AI arms race.
What is quantum computing and qubits?
Everything typed in classical computers, such as words and numbers, is translated into a binary code consisting of bits with a value of 0 (ground state) or 1 (excited state). However, a qubit leverages the principles of quantum mechanics to exist in both states simultaneously. For example, a qubit may have a 25% probability of having a value of 0 and a 75% probability of having a value of 1. This means that a single qubit can represent more information than a single classical bit.
As a result, quantum computers are able to process information in ways that are impossible for classical computers. They are able to solve problems that traditional computers cannot.
How are quantum computers different from supercomputers? Classical supercomputers excel at performing calculations at high speeds, relying on advanced architectures and acceleration techniques such as graphic processing units (GPUs) and multi-core processing. However, they are still bound by the constraints of classical computing principles and rely on logic gates such as AND, OR, XOR, and NOT gates to manipulate classical bits.
Quantum computers, on the other hand, use quantum gates such as H-gates and Pauli gates that are designed to process qubits and are reversible in nature. “Using these quantum gates, we can develop circuits and algorithms and solve problems that are otherwise impossible to solve,” explained Roy.
What is Google’s quantum computing chip Willow?
Google says its new state-of-the-art quantum computing chip is made at a facility in Santa Barbara, California, USA. The chip’s components include single- and two-quit gates, quiet reset, and readout that are engineered and integrated to ensure there is no delay between any two components as this can adversely affect system performance, the company said.
Errors are considered one of the biggest challenges in quantum computing as qubits, in superposition, rapidly exchange information with their environment and make it difficult to complete calculations. “Typically the more qubits you use, the more errors will appear, and the system becomes classical,” Google said.
However, with Willow, the company says it has successfully reduced errors when measuring the number of qubits processed by a quantum computer.
It tested an array of physical qubits, scaling from a 3 × 3 grid of encoded qubits, to a 5 × 5 grid, to a 7 × 7 grid. “Using our latest advances in quantum error correction, we were able to cut the error rate in half. In other words, we achieved an exponential reduction in the error rate,” Google said. First, quantum error correction in Willow happens in real-time, which is important because errors can ruin calculations if not corrected fast enough.
Additionally, the company said it put the Willow chip through the Random Circuit Sampling (RCS) benchmark test to measure its performance.
“Pioneered by our team and now widely used as a standard in the field, RCS is classically the most difficult benchmark that can be performed on a quantum computer today.
It tests whether a quantum computer can do something a classical computer cannot do,” Google said.
In the RCS benchmark evaluation, Google found that Willow was able to outperform one of the world’s most powerful classical supercomputers called Frontier.
“With 105 qubits, Willow now has best-in-class performance in the two system benchmarks discussed above: quantum error correction and random circuit sampling,” Google said.
How will Google’s quantum computing chip affect AI, encryption?
Giving an insight into the broader relationship between AI and quantum computing, Roy said that the domain of quantum AI includes developing AI algorithms and architectures with the advantages of quantum computing. “One of the key aspects of developing an AI model is to train it on large amounts of data. In that case, a quantum computer can be very helpful because it helps you compute the data faster,” he said.
Similarly, Google said it is exploring quantum algorithms to scale basic computational tasks for AI. It further highlighted that quantum computers will be able to collect training data for AI models that is currently inaccessible to classical computers.
When asked if Google will be able to level-up its AI play with quantum computing chips, Roy opined that the tech giant still has a long way to go. “The standard circuits that we use for standard AI models may not work. Some changes need to happen to ensure that AI models can work in quantum circuits, which is an active area of research,” he said.
A fully functional quantum computer might even have code-breaking capabilities that would render all forms of online encryption untrustworthy. RSA is a public-key encryption algorithm with many real-world applications such as digital certificates, digital signatures, virtual private networks, email encryption, and more. RSA is based on a problem called the discrete logarithm problem that is difficult for classical computers to solve.
However, in 1994, the American mathematician Peter Shore came up with an algorithm that showed that a quantum computer scaled to a certain capacity could break the discrete logarithm problem, and thus compromise the underlying cryptography of Bitcoin and other cryptocurrencies as well as any system. RSA encryption.
Does Google’s chip weaken RSA encryption? Not quite. Even with advances in error reduction and scalability, Willow is a 105-qubit chip and experts point out that it would take enough qubits to break RSA encryption.
“Estimates indicate that a quantum computer with approximately 13 million qubits would be required to achieve decryption within a 24-hour period when Bitcoin’s encryption is compromised,” said Kevin Rose, a tech entrepreneur and former senior product manager at Google.
However, even such a scenario has not been completely ruled out. With the increased focus on quantum computers in the past few years, researchers like Roy have been working to develop new, post-quantum algorithms that are secure against quantum computers.