Zapata: What You Should Know About Quantum Computing

As part of a recent interview, Enter Quantum asked Yudong Cao, chief technology officer and co-founder of quantum software company Zapata, what he wished companies knew about quantum computing. Here are his insights.

Enterprises should adopt quantum because classical computing is running out of steam and GPUs and CPUs started hitting bottlenecks 10 years ago. Now GPUs are all the rage but with the emergence of generative AI, GPUs may also become a saturated resource. GPU just shifts the curves so that on one side, you have computing mechanisms that are hitting bottlenecks, but on the other side, you have data that is accumulating exponentially.

That should paint a very stark picture for every enterprise to get ready for whatever computing resource they can get their hands on. I think enterprises will soon come into come to the realization – or maybe they're already in that reality now – that they have just a ton of data being accumulated that is just not being used.

Because there's a computing bottleneck, it's like you're drinking from a straw. The computing power is not there to leverage all that data, but at the same time that all the data is occupying space and, and costing money to maintain so it's a very easy ROI calculation to look for any kind of alternatives.

Quantum represents one of those alternatives to the conventional models of computing that is a lot more special purpose and tailored for these applications that are particularly compute intensive. I'm pretty sure you don't need GPUs to process emails or do any office applications but you do need GPUs for these advanced AI models in the future.

As those AI models advance along with more exotic computing resources, the exercise of getting ready for quantum also benefits the exercise of adopting all these other exotic computing modalities like FPGAs. I've seen special-purpose chips for doing mathematical operations that are unique to AI.