Neuromorphic computing, the underdog that bites at Quantum computing

This is not a spaceship — it’s a quantum computer.

Quantum computing seems to be the star candidate as a next-generation computing system, or so most headlines make us believe. The truth is that quantum computers have a lot of problems and limitations and even if they may have the ability to solve some problems at tremendous speeds, the problems they can solve are small in number and very specific. In the far-away future, when the technology will be mature and more quantum theory will help quantum computers be used better, they may prove as the ultimate computer. However, until that day arrives, another type of computer knocks at humanity’s door.

This Neuromorphic computer is researched as a candidate for a general problem-solving device that, with the use of artificial intelligence, may be used to solve almost every problem that can be solved using software deep learning agents. Put simply, this computer will enable us to simulate an inorganic brain hardware device, working similarly to the human brain. That means high computing power, highly parallelized (simultaneous) computations at low energy costs and unlike quantum computers, it won’t need extremely low temperatures to function.

Why not Quantum?

For today’s generation, a quantum computer is similar to the classical computer of our forefathers. If you look at the history of computers, you’ll notice that vacuum tubes were used to power computers the size of a room.

Today’s quantum computer is a sophisticated hardware arrangement the size of a room. Quantum processors require a very cold environment to function; we’re talking temperatures of nearly absolute zero [-273.15 C]. Because it operates at a subatomic level, this temperature aids in lowering the processor’s entropy. This entropy might cause errors in the processor’s readings, rendering it worthless.

It is, however, available as a service if you want to give it a shot. D-Wave Systems and IBM both provide a user interface for normal people with a small amount of quantum computing power. Don’t expect anything too amazing thou.

Google now has the most powerful CPU, a 72-qubit processor that is equivalent to a 272 conventional-bit processor. Scientists and developers are trying to figure out what these processors are capable of and what uses they might have.

It has the capability of accessing other universes, assisting us in the search for life on distant planets, and other sci-fi possibilities. This technology, however, will take decades to mature.

Why neuromorphic computers are more feasible for next-generation processing

Neuromorphic computing, as the name implies, is based on a model inspired by the brain’s operations. The brain is an interesting computing paradigm because, unlike most supercomputers, which take up entire rooms, the brain is small, fitting neatly into something the size of, well… your skull.

Your brain uses roughly 20 watts, whereas the Fugaku supercomputer requires 28 megawatts — or, to put it another way, a brain requires about 0.00007 percent of Fugaku’s power supply. While supercomputers, including quantum computers, require complex cooling systems, the brain is housed in a bone casing that maintains a constant temperature of 37°C.

Following Moore’s Law, chipmakers have been able to continue increasing the amount of computing power on a chip for a long time by packing more transistors onto these von Neumann computers. However, challenges with further downsizing transistors, their energy consumption, and the heat they emit suggest that this won’t be possible for much longer without a change in chip foundations.

Von Neumann architectures will become increasingly difficult to offer the gains in computing power that we require as time goes on.

Quantum computing and neuromorphic systems have both been offered as answers, although neuromorphic computing, also known as brain-inspired computing, is more likely to be commercialized first. This is because neuromorphic computing technology is considerably more advanced than quantum computing, and it does not have the same temperature limits or problematic physics theories.

What do you think?

Now it’s up to every one of us to decide whether we believe in neuromorphic or quantum computers more. Neuromorphic and analog computers, in my opinion, will be the first to enter our daily life, whereas quantum computers will remain a pet research topic for a long time.

More importantly, I believe that Neuromorphic computing is required for the development of a reliable quantum computer. To explore and build new technology, we need improved tools. Even though the Quantum computer is the one we desire, neuromorphic computing is the underdog we desperately need.