Introducing Azure Quantum: Quantum hardware unveiled by Microsoft
- Posted on March 8, 2021
- Estimated reading time 4 minutes
Microsoft recently unleashed Azure Quantum into public preview. In essence this new service allows us to scale up the services Microsoft have been developing in Quantum Inspired Optimization (using quantum methods to achieve speed up today on classical infrastructure) to cloud scale, but it also gives us access to something very exciting: real, working, qubit hardware consumable directly from Azure.
Trapped ion quantum computers use individual ions of atoms suspended in electromagnetic fields to hold qubit information in the electromagnetic state of each ion. The ions can be interacted with via electromagnetic forces and lasers to encode the qubits and then measure the changes in their states as the computations proceed. Trapped ion systems are shown to be the most accurate quantum hardware currently available.
Like all quantum systems, there is still a huge scalability issue that has yet to be tackled, meaning currently we only have public access to a 10 qubit Honeywell computer, and an 11 qubit IonQ computer. This is still somewhat behind the 16 qubit computer available from IBM since 2018 but could represent a more accurate technology than the superconducting system from IBM, and possibly a more scalable technology for the future.
For now, these are expensive resources, costing upwards of three thousand dollars for a single hour of compute time, and are relatively limited in qubit numbers due to the scalability. However, making these systems available on the cloud is a vital advancement. As these computers increase in qubits, they will be easily accessible for increasingly complex problem solving using the same code that is being written today.
It may come as a surprise that Microsoft are using external providers to quench today’s quantum thirst, especially since Microsoft have invested heavily to create an entire ecosystem in Azure built around quantum computing and QIO. It needs to be noted that these are relatively small quantum computers that we are being given access to.
Microsoft continues to innovate and make scientific gains on its own topographical quantum computing technology behind the scenes, hopefully a more promising long term bet for the future of quantum computing. It seems for now, however, that this technology may not be ready for consumption. Partnering allows Microsoft to test the Azure quantum ecosystem as well as build skills around Q# and train quantum software engineers of the future without the need to perfect public access to their own experimental hardware.
This is far from a quantum leap computationally, but it is a huge step in the right direction in terms of infrastructure, cloud services, and QIO. Running quantum ready algorithms on real quantum hardware gives significant opportunity to test and validate our understanding of how real quantum computers react to our solutions. This may include measuring performance against expectations, delving into qubit errors caused by the real quantum interactions, or importantly, allowing us to streamline the algorithms and processes needed to run these computations. All of this can position us to be ready to hit the ground running as more and more powerful hardware systems come online in the years to come, no matter which hardware technology or provider goes on to win the quantum race.