Working with real quantum computers just got easier for experts in chemistry, artificial intelligence and optimization. Building on QISKit, our open source quantum information science kit for software development, we’ve released AQUA – Algorithms and circuits for QUantum Applications. This new open source software allows classical computer applications to send complex operations to be run on quantum computers, over the cloud.
Let me start by explaining the quantum software stack, and where QISKit and AQUA fit. At the lowest level is the hardware where the qubits sit at the very cold temperature of 15 mK. The qubits receive microwave pulse signals for a calculation, which have been translated and converted from OpenQASM, IBM Q’s low-level assembly language, by QISKit. Users of the free IBM Q Experience 16 qubit system can write programs directly in OpenQASM, but it’s easier to use libraries in higher level languages. That’s where QISKit comes in. It’s a front-end interface that works with Python (read IBM Fellow Dr. Jay Gambetta’s article about QISKit’s recent upgrade).
QISKit alone requires developer skills. Running experiments on the IBM Q Experience means understanding how to write a program, or using someone else’s program such as those in QISKit’s github repository. So far, this approach has succeeded: More than 85,000 users have run more than four million experiments and published 80 research papers based on experiments run on the system. But we were missing the contributions of domain experts – until QISKit AQUA.
A classic-quantum hybrid environment
QISKit AQUA is a library of quantum algorithms that accomplishes two things:
It allows domain experts unfamiliar with quantum computing to access to IBM Q quantum computers via the classical applications they’re used to using, or via individual, existing, domain-specific algorithms.
It also allows researchers and developers to contribute new algorithms to QISKit AQUA’s open source domain libraries.
Beginning with chemistry, artificial intelligence and optimization, experts in these fields can begin to tap into QISKit AQUA as a new component to the applications they use in their research. They can do all this without fully understanding the complex quantum computation happening at the deeper levels of the software stack, or at the hardware level.
Today, QISKit AQUA Chemistry libraries support classical applications such as Gaussian, PSI4, PySCF and PyQuante. Users can work with one of these apps in ACQUA to run a specific execution on IBM Q quantum hardware or simulators – versus within the classical software – to examine the results a quantum computer produces. The release also includes QISKit ACQUA Artificial Intelligence and QISKit AQUA Optimization algorithms and will include interfaces to particular domain-specific applications in the near future.
Many working chemists today rely on software packages to do very specific computations related to their particular field of interest. We hope that chemists will look at how QISKit AQUA can extend the power of their familiar applications into the quantum realm. We’ve made it easy to download and install QISKit AQUA to start learning and experimenting.
Sixteen or twenty of today’s approximate universal qubits do not provide a specific advantage over classical computers just yet. IBM scientists even proved that 49 and 56 qubits can be simulated on supercomputers for particular problems. So, why create this hybrid classical-quantum environment now?
Experts can continue using their familiar domain-specific applications, which do much more than measure ground state energy of different molecules that we show in the demo. They can also take advantage of increases in quantum computing’s power as quantum volume improves, and the number of applications grows in QISKit AQUA’s respective libraries. QISKit AQUA advances industry-, academic- and research-wide collaboration to prepare for a world where classical and quantum computers work together to better solve computationally complex problems. The future of computing is hybrid and QISKit AQUA is a big step toward making that a reality.
A team of researchers from IBM Research AI and AI Horizons Network-partner the University of Michigan published the papers “A Large-Scale Corpus for Conversation Disentanglement” and “Learning End-to-End Goal-Oriented Dialog with Maximal User Task Success and Minimal Human Agent Use” at ACL 2019. This work address two main challenges in building enterprise AI assistants.