Smooth Axis in the Quantum Software Industry

We’ve noticed some subtle pivots in the quantum software industry over the last year or so. Many quantum software companies were founded as “pure” quantum software companies and demonstrated at the time that working on classical-based solutions would shift the focus of their efforts. But now, many of them are introducing classical based solutions, sometimes calling them quantum inspired, as part of their product portfolio. Some recent examples include QC Ware with them Prometheum Quantum Chemistry ApplicationsSandboxAQ is growing quantum simulation and optimization solutions which will run on GPU (Graphics Processing Unit) or TPU (Tensor Processing Unit), Zapata promotes their use of generative AIStrangeworks indicated they will be introduced soon a new tool that will take advantage of artificial intelligence technology, 1Qbit develops a ‘black box’ reinforcement learning method for simulated annealing (SA), and others. And to clarify, this is not a hybrid classical/quantum algorithm, but rather a pure classical, albeit innovative classical, solution.

It’s a pragmatic change for these companies and it makes perfect sense. End users usually don’t care how their problem is solved as long as it does. If a classics based solution can get there first it will be their preference. For software companies, this will generate revenue faster and make their investors happier. And in the meantime, they’ll be able to forge relationships with their end users and better understand the problems they’re trying to solve. In many cases, these companies are working on algorithm-driven solutions such as tensor networks that can be first deployed on GPUs and then updated relatively easily to run on true quantum computers once more powerful machines become available.

One thing to note is that while companies use many classic-based solutions today, many of them are not using the most modern or efficient solutions that may have been developed years ago. There are probably many organizations that are still trying to solve various optimization problems using Excel spreadsheets. So moving it to us, the best and most efficient modern classical algorithms still provide significant value. So while we wait for more powerful quantum computers to become available and allow pure quantum-based solutions or quantum/classical hybrids to become feasible, this is a good approach.

April 29, 2023