[ad_1]
A team of electrical and computer engineers from Rochester has developed a simple computing device that they believe could help solve problems of optimizing military logistics in complex battles in the future. (graphics by Getty Images)
Rochester researchers are developing new Ising machines with federal support for research and development funding from DARPA.
“Why Russia’s Army Is Stuck in Logistics.” “Allies Fail to Agree on Sending Tanks to Ukraine.”
These recent headlines highlight the importance of logistics in war. What weapons and supplies are needed? In what quantity? And just as important, what is the most cost-effective way to get these supplies to the right places and at the right time to soldiers on the front lines spread hundreds, even thousands, of miles away?
A team of electrical and computer engineers at the University of Rochester believe their invention – a simple computing device like no other – could help solve problems of optimizing military logistics in complex battles in the future.
To that end, the Defense Advanced Research Projects Agency (DARPA), the research and development agency of the United States Department of Defense, recently awarded researchers a quantum-inspired classical computing grant—which could amount to 6, $1 million over five years – to develop two new Ising machines.
Ising machines that outperform quantum computers?
Named after the German physicist Ernst Ising, the Ising model describes how atoms in natural magnets or spin glasses adopt one of two values—spin up or spin down—to arrange themselves in the lowest energy state. Ising machines are designed to mimic and further refine this process to find optimal solutions to so-called combinatorial optimization problems that involve a large number of competing alternatives. Moreover, in such problems, the number of possible solutions increases exponentially as the number of independent variables increases.
The machines will be tested on combinatorial optimization problems posed, for example, when delivering tens, even hundreds, of combat units, says the project’s principal investigator Michael Huang, professor of electrical and computer engineering. They can also be used in many commercial applications, such as finding efficient routes for packet delivery, generating test patterns for chip error detection, and efficient error correction for 5G wireless radios.
Compared to conventional computers, “our devices have a much simpler architecture,” says Huang. “It can only solve these kinds of optimization problems. We can’t do Zoom calls on it. So it is a special purpose machine. But he’s extremely good at what he does.”
The devices are still in the early stages of development by the Rochester team, which includes Gonzalo Mateos, Zeljko Ignatowicz, Chiang Lin, Selcuk Kose and Hui Wu, also faculty members in the Department of Electrical and Computer Engineering. One of the devices, however, underwent extensive simulations that showed it would solve moderate-sized optimization problems several orders of magnitude faster than conventional computers — and with much less energy, Huang says.
Compared to already existing $15 million quantum annealing devices (a type of quantum computer), the simulated device will be “insanely cheap” to build and will be compatible with already existing CMOS (complementary metal oxide semiconductor) integrated circuits, he adds.
Huang is confident that the device will surpass even quantum computers in speed and power in solving optimization problems “at least in the current NISQ [noisy intermediate-scale quantum] era,” he says. Also, unlike quantum computers, which require cryogenic conditions, both devices his team is proposing will work at room temperature.
Using the natural laws of physics instead of conventional calculations
Many events in nature — for example, an object falling to earth — can be modeled by writing and solving differential equations, according to Huang. “This suggests that nature itself is somehow effectively solving these differential equations,” he says.
Ising machines attempt to directly use the computations performed by nature. This is sometimes called “nature-based computing” or “physical computing”.
To date, various Ising machines have been developed, including optics-based quantum designs. However, they all have important practical limitations imposed by their design, Huang says. Some only have possibilities to connect close neighbors in the different rotations, also called nodes. Others use conventional calculations to emulate binding, thereby losing efficiency; however, others require elements that are difficult to integrate into a chip.
The first Ising machine the Rochester team is developing is based on the BRIM (Bistable Resistively-Coupled Ising Machine) architecture. Developed by Huang and Ignjatovic, it will provide connectivity across all nodes without emulation while using elements that are easy to integrate on a chip. “I call it the world’s first Ising machine without major flaws,” says Huang.
To date, the researchers have simulated the device in tens of thousands of nodes. “When you get to very large scale—hundreds of thousands of nodes or more—we can run into a potential latency issue that can limit its performance,” Huang says.
As a more futuristic alternative, the team will also develop a second optics-based Ising machine, which Huang and Lin began researching nearly six years ago. Higher speed machines based on optics have the potential to address critical path delay issues. However, large-scale photonic systems are generally “more challenging to build simply because they are not yet supported by mature manufacturing technology,” Huang says. “So it’s going to be more of a high-risk, long-term decision.”
A third part of the grant, led by Huang and Mateos, involves developing a software/hardware code design to efficiently map problems to the two hardware platforms. And if fully funded, the project will support the equivalent of 12 PhD students, or six PhD students and three postdocs in the researchers’ labs.
Read more
Category: Science and Technology
[ad_2]