IBM has outlined its quantum computing development roadmap that will begin with the release of the Qiskit Runtime open-source software in 2021...
Quantum computing is at a pivotal point. The decisive quantum leap could be coming. IBM has outlined its quantum computing development roadmap that will begin with the release of the Qiskit Runtime open-source software in 2021.
In an interview with EE Times Europe, Bob Sutor, vice president of Quantum Ecosystem Development at IBM, pointed out that currently no one is yet using quantum computers in production. The challenge will be to make this new environment more and more accessible and allow companies and developers to experiment with the release of applications. Besides software aspects, the increasingly efficient hardware with a decidedly high qubit count will pave the way for commercial applications. In any case, in the near future, quantum computers will not replace classical computers. They will work together.
In September 2020, IBM highlighted how they took the bold step of releasing a hardware roadmap that shows a clear path to over 1,000 qubits and identified challenges along the way. This time around, there will be the open-source community and the ability to mobilize developers around the world, plus cloud-native deployment to break down critical challenges and democratize access to this new technology as quickly as possible.
IBM and open-source developers will work to optimize the foundation of the stack. Simultaneously, other kernel developers are deploying high-performance quantum circuits with increasingly sophisticated performance mechanisms, paving the way to develop models for chemistry, physics, biology, machine learning, optimization, or even finance.
“Our new development roadmap provides new opportunities for collaboration. We are increasing the variety of circuits and the ability of our systems to run more circuits faster. The roadmap is pointing out beyond 2023, when we will get above 1,000 qubits. Then, we can seriously address things like error correction, and look at Quantum Advantage, which is the point where quantum systems together with classical systems can do much better than just classical systems by themselves,” said Sutor.
He added, “the ability to reliably manage the operation and control of different quantum systems was not possible just a few years ago. Today, we can increase the number of qubits, thanks to extraordinary efforts in science and engineering.”
The implementation of the roadmap will be done in stages. It will include the implementation of high-performance quantum software and hardware. The development of new quantum algorithms will build on the foundation of innovative quantum circuits, thus the development of programs or applications. On top of the algorithms, complex programs and models will be used in various industry fields.
Hardware for quantum computing
It will take another few years for us to see commercial applications in production. In the meantime, many leading companies are developing ambitious quantum computing programs. A number of startups are investing in quantum computing especially in Error Quantum Correction and Cryptography.
In 2001, IBM developed the first 7-qubit quantum processor. In 2016, IBM released the first quantum system openly available on the cloud. IBM now has approximately 20 computers in the cloud available to users for free.
IBM plans to reach 127 qubit quantum computers by the end of 2021, 433 qubits before the end of next year, and more than 1,000 qubits by 2023.
The company’s newly released processor, IBM Quantum Hummingbird, is 65 qubits. Next year, it will be IBM Quantum Eagle, at 127 qubits, while in 2023 it expects to release IBM Quantum Condor, a processor at 1,121 qubits. They expect eventual quantum computing applications in the field of neural networks moving to the realization of economic models, from the study of personalized drugs to simulations of complex chemical reactions up to many other possibilities such as the optimization of mathematical problems and some parts of artificial intelligence.
The IBM Quantum Eagle chip will feature several upgrades to reduce qubit errors and will continue to lay the groundwork for scaling the number of qubits that work together as logical qubits. IBM pointed out that, with the Eagle processor, real-time classical compute capabilities will be introduced for execution of a broader family of quantum circuits and codes.
In quantum computers, the bits, i.e., the basic information unit of classical computers, are replaced by the so-called “qubits”, quantum bits, which are able to cope with enormously complex problems thanks to a greater possibility of encoding information. Problems that are largely out of reach for normal computers. Like classical computers that include logic gates and circuits, a quantum computer uses quantum circuits composing elementary quantum logic gates.
Quantum computers use three concepts. The first is the “quantum superposition”, the idea behind Schrödinger’s living and dead cat. Unlike the classic bits, that can only have two states – one or zero -, the “qubits” states can be a combination of both. The second is “entanglement”. It correlates quantum particles together through time and space. The third is interference, where we can design algorithms so that the “right answers” are more likely to appear than the wrong.
Open source for quantum computing
IBM has specified that quantum computers will be particularly useful in certain areas: Life sciences, chemistry, and artificial intelligence. In these areas, the open-source software Qiskit will be able to accelerate certain computational tasks. This year, IBM will release the Qiskit Runtime – an execution environment that increases the capacity to run more circuits at a much faster rate than ever before. This increases the capacity of a quantum computer to do more work.
And just to make the most of Qiskit’s capabilities, IBM is expanding the pool of developers involved in the project. “We’ve trained thousands people to use Qiskit, and the software has been downloaded by developers more than half a million times,” said Sutor.
He added, “With this roadmap, we are talking about how to improve our systems in three different ways. The first is quality. How well will circuits do what they are supposed to do? This is the Quantum Volume discussion. It includes things like error mitigation and reduction of noise. The second thing is capacity. How fast are your machines? We aim to run circuits 100 times faster by the end of the year. The increased capacity gives me more time to run more circuits. And the third type is variety, that is, getting the benefits of quantum with some of the most interesting features of classical coding.”
Qiskit provides a set of code tools for quantum circuit-level programs, offering execution and management on remote access back ends. IBM is making the functions very simple even for those who are not experts in quantum theory or quantum mechanics, which is the basis of a quantum computer. The goal is to offer a wider variety of circuits, allowing users to tackle problems unsolvable with classical computers. Software tools such as OpenQASM 3 will soon offer quantum kernel developers the ability to run dynamic circuits – those that incorporate both classical and quantum instructions that must be executed within the coherence time of qubits – by 2022. Sutor pointed out that in 2023 there will be new advanced control systems to manipulate large amounts of qubits to provide the full benefit of quantum computing.