Curbing the power consumption of data centers using quantum computers

Data centers are at the heart of all things Internet. Whether you’re modeling complex, biological systems or liking someone’s social media post, data centers play an essential part.

But our increasing dependence on information technology comes at a price: increasing dependence on energy. Most of this growth comes from the industrial sector, which accounts for more than double the energy consumption of the residential sector. Datacenter sustainability depends on the actions of large companies even more than it does on consumer behavior. 

In 2018, global data centers accounted for over 2% of the global energy use - a staggering figure. Today, there are over 6000 data centers around the world. Industries that use data centers are responsible for great innovations like drug discovery, machine learning, weather forecasting, traffic optimization, and more. As additional questions are answered, many new questions are revealed, increasing the need for better, faster, and more advanced technology.

If we continue down this path of uninhibited growth, data center use will continue to increase, and with it the use of energy to power then. A bleak future could be within reach.

We have two options: stop innovation and all things energy-dependent, or continue the growth, but do it in an energy-conscious fashion. We can probably unanimously agree on the latter being the right move. Quantum computing is one emerging solution to drive and even accelerate innovation while reducing energy use is quantum computing.

When complex problems are appropriately translated into quantum computing code, quantum computers can deliver much higher performance yet consume much less energy.

Google’s “quantum supremacy” experiment of 2019 compared the execution time of the algorithm on quantum and classical computers, but an interesting byproduct of that experiment was a comparison of the energy usage.. A study from NASA, Google, and Oak Ridge National Lab compared the energy consumption of a supercomputer with that of a quantum computer solving the same problem. The quantum computer required a tiny fraction, a mere 0.002% of the energy of the supercomputer. It might be that through quantum computing, we can forge a more sustainable future.

Here’s why quantum computers are more efficient: a conventional data center computer may use billions of transistors, but with a quantum computer, you have hundreds, thousands, or eventually millions. Depending on the particular implementation, a qubit might be just a single atom. That means that you only need enough energy to excite, or move around, millions of atoms instead of billions of transistors. Additionally, when a problem is properly coded, quantum computers can analyze massive data sets in parallel; whereas classical computers need to analyze them serially.

It must be mentioned that quantum computers are not going to replace classical computers in all, or even in most, data center applications. But specific compute tasks like optimizations, chemical simulations and some flavors of machine learning do appear to be promising applications for quantum.

How do we get there? It seems that every month, larger quantum computers are announced. The quantum hardware race is in full swing.

But we must make sure that the software is keeping up. Are we ready to code the quantum computers that will outperform classical ones? The learning curve in designing quantum algorithms is steep, and coding practices that work for small quantum computers are inadequate for larger machines. There is an acute need for software that can keep up with the improving hardware.

New software platforms like Classiq offer an environment for designing complex quantum circuits that would be impossible to create by hand and help quantum computers reach their potential, yet make quantum computing more accessible to a wider range of non-quantum experts.

With the right software, quantum computers can help the planet not only by developing better EV batteries or greener ways of producing fertilizers, but also by curbing our energy consumption.