Power/Performance Bits: Nov. 8

Molecular memristor; multi-beam mmWave.


Molecular memristor
Researchers from National University of Singapore, Indian Association for the Cultivation of Science, University of Limerick, Texas A&M University, and Hewlett Packard Enterprise discovered a molecular memristor for brain-inspired computing.

The molecule uses natural asymmetry in its metal-organic bonds to switch between different states, which allows it to perform ultra-fast decision-making.

“In the new device, everything is done in one place, so there is no need to keep reading or moving information around,” said Damien Thompson, professor in physics at University of Limerick. “This removes the ‘von Neumann bottleneck’, a problem that has plagued computing from the very beginning and still hampers technology development. The new molecular circuitry means the computer-processing unit no longer has to fetch data for every operation it performs, and this saves enormously on time and energy costs.”

The researchers created an electrical circuit consisting a 40-nanometer layer of molecular film from the chemical family of phenyl azo pyridines sandwiched between a top layer of gold, and a bottom layer of gold-infused nanodisc and indium tin oxide. They observed an unprecedented current-voltage profile upon applying a negative voltage to the device. Unlike conventional metal-oxide memristors that are switched on and off at only one fixed voltage, these organic molecular devices could switch between on-off states at several discrete sequential voltages.

“Similar to the flexibility and adaptability of connections in the human brain, our memory device can be reconfigured on the fly for different computational tasks by simply changing applied voltages. Furthermore, like how nerve cells can store memories, the same device can also retain information for future retrieval and processing,” said Sreetosh Goswami, research fellow from the Department of Physics at NUS.

“These molecules are like electron sponges that can offer as many as six electron transfers resulting in five different molecular states. The interconnectivity between these states is the key behind the device’s reconfigurability,” added Sreebrata Goswami, a senior research scientist at NUS.

The team used the molecular memory devices to run programs for different real-world computational tasks. As a proof of concept, the team demonstrated that their technology could perform complex computations in a single step and could be reprogrammed to perform another task. They posit that an individual molecular memory device could perform the same computational functions as thousands of transistors.

Thompson added, “We are excited about the possibilities because the devices show all the hallmarks of brain computing. First, a huge number of tiny, identical molecular processors are networked together and work in parallel. More importantly, they show both redundancy and reconfigurability, which means the device can solve problems even if the individual components do not all work perfectly all the time or in the exact same way every time.”

Multi-beam mmWave
Engineers from the University of California San Diego propose a way to make millimeter wave 5G more reliable. mmWave signals are easily blocked by objects and have limited rage. Instead of the current approach of sending one mmWave beam between a base station and receiver, the researchers suggest splitting it into multiple beams. Each beam would take a different path from the base station to the receiver.

The researchers created a system for splitting mmWave beams and tested it inside an office and outside a building on campus. The system provided a high throughput connection (up to 800 Mbps) with 100% reliability. The signal didn’t drop or lose strength as the user moved around obstacles like desks, walls, and outdoor sculptures. In outdoor tests, the system provided connectivity up to 80 meters (262 feet) away.

The system relies on a set of new algorithms. One algorithm first instructs the base station to split the beam into multiple paths. Some of these paths take a direct shot from the base station and the receiver, while and some paths take an indirect route to get to the receiver. In these indirect routes, the beams bounce off reflectors, which are surfaces in the environment that reflect millimeter waves like glass, metal, concrete, or drywall. The algorithm then learns which are the best paths in the given environment. It then optimizes the angle, phase, and power of each beam so that when they arrive at the receiver, they combine constructively to create a high quality and high throughput signal.

“You would think that splitting the beam would reduce the throughput or quality of the signal,” said Dinesh Bharadia, a professor of electrical and computer engineering at the UC San Diego Jacobs School of Engineering. “But with the way that we’ve designed our algorithms, it turns out mathematically that our multi-beam system gives you a higher throughput while transmitting the same amount of power overall as a single-beam system.”

Another algorithm maintains the connection when a user moves around and when another user steps in the way by continuously tracking the user’s movements and realigning the beam parameters.

“You don’t need any new hardware to do this,” said Ish Jain, an electrical and computer engineering Ph.D. student at UC San Diego. “Our algorithms are all compliant with current 5G protocols.”

The team is working to scale the system to accommodate multiple users.

Leave a Reply

(Note: This name will be displayed publicly)