For the past 50 years the capabilities of computers have been dictated by Moore’s Law, which states that the processing power of computer chips doubles every 18 months as manufacturers devise ways to fit more transistors on smaller bits of silicon. But what if instead of doubling, that computing power could be increased by a thousand fold? What could be done with all that extra power?
Orlando Hernandez, associate professor of computer engineering at the College of New Jersey, sees a leap of that size in the near future as a real possibility thanks to advances in a technology known as quantum computing. And if a commercial quantum computer is ever made, the effects could be revolutionary.
Hernandez is a panelist at the New Jersey Technology Council’s discussion of emerging tech trends, which will take place Thursday, February 21, from 4 to 7 p.m. at Maestro Technologies at 1 West State Street in Trenton. Hernandez will be on a panel of educational and government officials including Andrew Lowman, dean of Rowan’s school of engineering, Brian Sabina of the NJ Economic Development Authority, and assemblyman Andrew Zwicker, who chairs the science, innovation, and technology committee and holds a PhD in physics from Johns Hopkins University. A corporate panel will feature Sean Wohtman, customer engineer for Google, Chris Sullens of Central Reach, Walter Willinger, chief scientist of NIKSUN, and Steve Brittin of ORS Partners. Tickets are $40, $20 for members. For more information visit www.njtc.org.
Hernandez says the acceleration of silicon chip technology is slowing down due to manufacturing and power consumption constraints. Quantum computing offers an alternative approach that would use lower power and have higher density and computing power.
Traditional computing is based on electrical circuits that are either turned on or turned off at any given time and uses “bits” of information that are either a 1 (on) or 0 (off). Quantum computing takes advantage of phenomena of quantum mechanics at the atomic level, in which particles can be in more than one place at the same time. Because of this property, each “qubit” in a quantum computer can store vastly more information than just a one or a zero.
Multiple Silicon Valley companies — Google, for one — have created experimental quantum computers that demonstrate the principles involved, but they are not yet practically useful. Hernandez thinks that if there is a breakthrough and someone creates a commercial quantum computer, the effects would be transformative.
“Computing performance would move forward beyond the devices we can make right now out of silicon. We could use more sophisticated types of algorithms that artificial intelligence and machine learning are going to demand,” he says.
For instance, artificial intelligence could begin to use “brain-inspired algorithms.” The human brain is made up of about 6 billion interconnected neurons and uses very little power, but it uses electrochemical processes that are very slow. A quantum computer could simulate a similar large amount of interconnections, but the connections could work at lighting speed.
“We could have brain-like capabilities, which would allow us to do very sophisticated types of algorithms,” Hernandez says.
Hernandez says one of the biggest obstacles to making practical quantum computers a reality is manufacturing. Silicon transistors have a vast global infrastructure dedicated to making them that has been built over the last 70 years along with the knowledge to go along with it. The manufacturing methods have become highly efficient. Every sheet of transistors can be cut into hundreds of identical processors. Could quantum computers ever be made with such efficiency and reliability? It’s an open question. “It’s one thing to make a few, but can you make millions?” Hernandez says.
Meanwhile, devices powered by conventional computers are becoming ever more advanced. Hernandez believes that true self-driving cars will hit the market sooner than skeptics predict. In the long run, he sees the potential for the introduction of self-driving vehicles to totally remake the road infrastructure. For example, traffic lights might some day disappear as self-driving cars develop the capability to communicate with one another and organize safe and efficient passage through intersections.
“You can envision not just autonomous vehicles, but a whole infrastructure of traffic management,” he says. “It’s just going to be more a question of is society going to accept it? I worry about societal issues such as regulation and litigation.”
Hernandez predicts that we will soon see expanded uses for aerial autonomous vehicles. “Companies that are at the forefront of technology, like Amazon, have huge interest in autonomous systems for delivering packages and things,” he says. Hernandez says the biggest bottleneck in the product-to-consumer supply chain is the delivery. Delivery, he says, “is a big commercial area” with a good business case for using drones.
“People are always going to need to get things, whether it’s food, or cleaning items, or devices, and I think that this industry is the biggest opportunity for autonomous systems,” he says.
The supply chain could be even more efficient, Hernandez says, if 3D printers improve to the point where it becomes practical for people to print objects in their own homes, eliminating more delivery vehicle trips.
Hernandez sees drone technology advancing rapidly by land, air, and sea. Seaborne drones have few civilian applications so far but are incredibly useful for military, safety, and search-and-rescue applications.
When it comes to military drones, Hernandez is on the side of those who believe that military robots should never be able to make the decision to kill on their own. “I would think from an ethical and political perspective, ultimately the human, and ideally several levels of human decision making checkpoints should be involved,” he says.
Hernandez was born in Cuba but grew up in Tampa, Florida, where his father was a baker. He loved to tinker with electrical devices from a young age, and this interest extended to personal computers, which became popular while he was in college in the 1980s. “I basically got one right away and used it for school,” he says. “I started studying the architecture of processors and got into chip design.”
He earned bachelor’s and master’s degrees from the University of South Florida and a PhD from Southern Methodist University, all in electrical engineering. His first job out of college was with Texas Instruments. He eventually worked on machine vision systems for TI and earned a doctorate in that field. His dissertation was on how to use mathematical models to allow computers to detect scenery and discern important details.
Today he is char of electrical and computer engineering at TCNJ.