Periodic Reporting for period 3 - DA QC (Design Automation for Quantum Computing)
Periodo di rendicontazione: 2023-09-01 al 2025-02-28
Quantum computing covers the development of applications that exploit quantum-mechanical effects, such as superposition and entanglement, in order to solve certain problems significantly faster (in the best case, exponentially faster) than classical computers. While considered “dreams of the future” for decades, this technology is currently maturing from an “academic vision” implemented in laboratory experiments towards a practical reality supported by compute centers.
This leads to an increasing complexity in the design of corresponding applications and systems that will soon not be manageable manually anymore, demanding dedicated design automation solutions. In the classical realm (i.e. for electronic circuits and systems), sophisticated design tools are taken for granted today and constitute a main reason for the penetration of electronic devices into almost all parts of our daily lives. In contrast, the development of design methods for quantum computing is still in its infancy. The three main reasons for this are:
Complexity: Even presumably simple tasks, such as the logic simulation of a circuit, suddenly yield exponential complexity. The experience of the design automation community with problems of similar complexity may lead to efficient solutions. But most of the experts in quantum computing do not have a background in design automation, and, hence, this potential is left untapped.
Terminology and Formalizations: Vice versa, experts in design automation often do not have a sufficient background in quantum computing. Additionally, since terminologies as well as formalizations are often unclear or ambiguous to design automation experts, “wrong problems” are frequently addressed or inappropriate models, faulty assumptions, or misleading cost metrics are used.
Interdisciplinarity: A closer interaction between both communities would help address many of these problems. But bridging this gap requires lots of resources and time to exchange with experts and stakeholders in these fields, which is complicated by their different languages as well as cultures.
Overall, the little interaction and exchange between experts in design automation and quantum computing are the main reasons why many relevant problems in the design of quantum computing have not been sufficiently considered yet. As a result, we may end up in a situation where we have powerful quantum computers but hardly any proper means to actually use them.
This leads to an increasing complexity in the design of corresponding applications and systems that will soon not be manageable manually anymore, demanding dedicated design automation solutions. In the classical realm (i.e. for electronic circuits and systems), sophisticated design tools are taken for granted today and constitute a main reason for the penetration of electronic devices into almost all parts of our daily lives. In contrast, the development of design methods for quantum computing is still in its infancy. The three main reasons for this are:
Complexity: Even presumably simple tasks, such as the logic simulation of a circuit, suddenly yield exponential complexity. The experience of the design automation community with problems of similar complexity may lead to efficient solutions. But most of the experts in quantum computing do not have a background in design automation, and, hence, this potential is left untapped.
Terminology and Formalizations: Vice versa, experts in design automation often do not have a sufficient background in quantum computing. Additionally, since terminologies as well as formalizations are often unclear or ambiguous to design automation experts, “wrong problems” are frequently addressed or inappropriate models, faulty assumptions, or misleading cost metrics are used.
Interdisciplinarity: A closer interaction between both communities would help address many of these problems. But bridging this gap requires lots of resources and time to exchange with experts and stakeholders in these fields, which is complicated by their different languages as well as cultures.
Overall, the little interaction and exchange between experts in design automation and quantum computing are the main reasons why many relevant problems in the design of quantum computing have not been sufficiently considered yet. As a result, we may end up in a situation where we have powerful quantum computers but hardly any proper means to actually use them.
During the first 3.5 years of the project, significant progress has been achieved towards the goal of building a bridge between the design automation community and the quantum computing community. To this end, numerous outreach activities as well as a series of technical exchange meetings have been conducted.
These have connected us -- as design automation experts from computer science -- to the experimentalists developing hardware platforms, physicists and mathematicians developing theoretical models, and domain experts acting as potential users.
As a result of these continued activities, several scientific methods have been developed that explicitly bring design automation expertise to quantum computing problems. The methods that were developed cover the whole quantum computing software stack, from developing relevant applications to the quantum computing hardware itself. They include solutions that help end users, improve quantum circuit simulations, offer compilation methods from high-level applications to low-level implementations, and allow one to verify the results.
In the recent period in particular we have made strong advancements in the areas of applications and application-aware tools, as well as quantum error-correction as one of the most timely and important topics in quantum error correction.
Moreover, we have led joint efforts between high-performance computing experts and other stakeholders in the Munich Quantum Valley, to develop the Munich Quantum Software Stack, an exposition of which can be found at https://www.munich-quantum-valley.de/de/forschung/forschungsbereiche/mqss(si apre in una nuova finestra).
Besides numerous scientific publications, all the developed methods have been made publicly available as open-source implementations through the Munich Quantum Toolkit (MQT)—an ever growing collection of design automation tools and software for quantum computing. What started out as set of independent tools has increasingly gained traction and a professional level. Finally, we are proud to announce that we have founded a spin-off, the Munich Quantum Software Company, whose goal is to further improve the quality of the tools developed for research methods to an industry grade standard. More information about the MQT is available at https://www.cda.cit.tum.de/research/quantum/mqt/(si apre in una nuova finestra) and the website of the company is https://munichquantum.software/(si apre in una nuova finestra).
The Munich Quantum Software Forum, an exchange and dissemination event we organized in October 2023 has had its second occurrence again in 2024.
We were blown away by the continued and even increased interest in our conference and were happy to welcome over 200 participants from industry -- both established companies like Google, IBM, Intel, NVIDIA and quantum startups like IQM, Planqc, Xanadu) and academia.
A video available at https://youtu.be/DRbCLtQWTls(si apre in una nuova finestra) provides a brief summary of the event and all talks have been recorded to live up to a modern conference standard. The videos of all talks that also demonstrate the diversity and interdisciplinarity of the participants are available at https://www.youtube.com/@CDA-TUM/videos(si apre in una nuova finestra) where also the video summary of the previous edition, MQSF 2023 can be found.
These have connected us -- as design automation experts from computer science -- to the experimentalists developing hardware platforms, physicists and mathematicians developing theoretical models, and domain experts acting as potential users.
As a result of these continued activities, several scientific methods have been developed that explicitly bring design automation expertise to quantum computing problems. The methods that were developed cover the whole quantum computing software stack, from developing relevant applications to the quantum computing hardware itself. They include solutions that help end users, improve quantum circuit simulations, offer compilation methods from high-level applications to low-level implementations, and allow one to verify the results.
In the recent period in particular we have made strong advancements in the areas of applications and application-aware tools, as well as quantum error-correction as one of the most timely and important topics in quantum error correction.
Moreover, we have led joint efforts between high-performance computing experts and other stakeholders in the Munich Quantum Valley, to develop the Munich Quantum Software Stack, an exposition of which can be found at https://www.munich-quantum-valley.de/de/forschung/forschungsbereiche/mqss(si apre in una nuova finestra).
Besides numerous scientific publications, all the developed methods have been made publicly available as open-source implementations through the Munich Quantum Toolkit (MQT)—an ever growing collection of design automation tools and software for quantum computing. What started out as set of independent tools has increasingly gained traction and a professional level. Finally, we are proud to announce that we have founded a spin-off, the Munich Quantum Software Company, whose goal is to further improve the quality of the tools developed for research methods to an industry grade standard. More information about the MQT is available at https://www.cda.cit.tum.de/research/quantum/mqt/(si apre in una nuova finestra) and the website of the company is https://munichquantum.software/(si apre in una nuova finestra).
The Munich Quantum Software Forum, an exchange and dissemination event we organized in October 2023 has had its second occurrence again in 2024.
We were blown away by the continued and even increased interest in our conference and were happy to welcome over 200 participants from industry -- both established companies like Google, IBM, Intel, NVIDIA and quantum startups like IQM, Planqc, Xanadu) and academia.
A video available at https://youtu.be/DRbCLtQWTls(si apre in una nuova finestra) provides a brief summary of the event and all talks have been recorded to live up to a modern conference standard. The videos of all talks that also demonstrate the diversity and interdisciplinarity of the participants are available at https://www.youtube.com/@CDA-TUM/videos(si apre in una nuova finestra) where also the video summary of the previous edition, MQSF 2023 can be found.
The results obtained in this project thus far showcase the (missed) potential. The corresponding outreach activities confirmed its demand. By that, we made the first important steps towards establishing design automation for quantum computing—a goal for which the current results provide a great basis and which shall be completed in the remaining duration of the project.
To this end and as a main contribution of the project, we created “bridges” to various disciplines and established numerous cooperations with stakeholders. By doing so, we address the main problem that, thus far, there has been far too little coordination between the design automation community and the quantum community, ensuring that the design approaches developed in this project address the “right problems” or succeed in reaching the end users. Our network of partners and contacts spans dozens of entities from industry and academia and goes beyond other projects and developments.
As one particular example, the project heavily works together with initiatives such as the Munich Quantum Valley (MQV)—a huge quantum computing initiative composed of over 300 researchers from various domains such as physics, engineering, computer science, domain experts, etc. that aims at developing a full quantum software stack in a holistic and full-stack manner. Obviously, the tools developed within this project perfectly address the needs of the quantum design tools that are part of this flow. Besides that, the methods developed in this project are also integrated into other huge software stacks.
Furthermore, we take an active role in “community building” by organizing events such as the Munich Quantum Software Forum mentioned above or special sessions, tutorials, panels, etc.—particularly for communities outside our core discipline. We additionally introduce these topics and the results from our project to other companies, cover them in advisory boards, governmental initiatives, and many more. Through all that, we significantly improve the state of the art by introducing design automation for quantum computing in various fields, working towards a common terminology, and taking on the role of an interpreter or mediator between varying disciplines, all of which are needed to make quantum computing a success.
To this end and as a main contribution of the project, we created “bridges” to various disciplines and established numerous cooperations with stakeholders. By doing so, we address the main problem that, thus far, there has been far too little coordination between the design automation community and the quantum community, ensuring that the design approaches developed in this project address the “right problems” or succeed in reaching the end users. Our network of partners and contacts spans dozens of entities from industry and academia and goes beyond other projects and developments.
As one particular example, the project heavily works together with initiatives such as the Munich Quantum Valley (MQV)—a huge quantum computing initiative composed of over 300 researchers from various domains such as physics, engineering, computer science, domain experts, etc. that aims at developing a full quantum software stack in a holistic and full-stack manner. Obviously, the tools developed within this project perfectly address the needs of the quantum design tools that are part of this flow. Besides that, the methods developed in this project are also integrated into other huge software stacks.
Furthermore, we take an active role in “community building” by organizing events such as the Munich Quantum Software Forum mentioned above or special sessions, tutorials, panels, etc.—particularly for communities outside our core discipline. We additionally introduce these topics and the results from our project to other companies, cover them in advisory boards, governmental initiatives, and many more. Through all that, we significantly improve the state of the art by introducing design automation for quantum computing in various fields, working towards a common terminology, and taking on the role of an interpreter or mediator between varying disciplines, all of which are needed to make quantum computing a success.