Manufacturing Bits: Nov. 13

Quantum memories
The University of Alberta has developed a new method for making quantum memories, paving the way for a next-generation quantum Internet.

Quantum memory is targeted for quantum networks and computers. In classical computing, the information is stored in bits, which can be either a “0” or “1”. In quantum computing, information is stored in quantum bits, or qubits, which can exist as a “0” or “1” or a combination of both. The superposition state enables a quantum computer to perform millions of calculations at once.

Quantum communications or networks follow the same basic idea. A quantum fiber link connects one location to another, which is supposedly impossible to hack. Traditional communication networks use public key cryptography. In contrast, quantum key distribution (QKD) uses quantum superposition states for unconditional security.

Quantum networks have been demonstrated over short distances. But long distance QKDs are limited, due to losses in the optical fibers, according to the Centre of Excellence for Quantum Computation and Communication Technology in Australia.

An optical quantum memory is used in a repeater device in a quantum network. It will help extend the range of quantum key distribution, according to the organization. Longer term, quantum computers will require quantum memory.

Still in R&D, quantum memories have been in the works for years. Today’s quantum memories are based on various coherent light–matter interaction schemes, but they are limited due to a host of technical challenges, according to the University of Alberta.

In response, researchers from the University of Alberta have devised a technique to overcome the challenges. Researchers have demonstrated a proof-of-concept storage device and signal processing capabilities in the quantum memory arena.

“We’ve developed a new way to store pulses of light—down to the single-photon level—in clouds of ultracold rubidium atoms, and to later retrieve them on demand by shining a ‘control’ pulse of light,” said Lindsay LeBlanc, assistant professor of physics and Canada Research Chair in Ultracold Gases for Quantum Simulation.

There are other innovations as well. “This scheme relies on dynamically controlled absorption of light via the ‘Autler–Townes effect,’ which mediates reversible transfer between photonic coherence and the collective ground-state coherence of the storage medium,” according to researchers in a journal called Nature Photonics.

Neuron circuits
HRL Laboratories has demonstrated a memristor that electronically mimics neurons in a brain.

HRL has demonstrated electronic neuron circuits that exhibit as many as 23 known behaviors of biological neurons. The organization also devised three classes of neuron activation.

These circuits were built using two nanoscale switches called active memristors, which is a type of resistive RAM. The memristors are based on a vanadium dioxide material, which is sandwiched between two metal layers. The combined material can switch between an insulating and a metallic phase.

HRL Laboratories has developed novel scalable and biomimetic electronic neurons built upon nanoscale vanadium dioxide active memristor devices. (© 2018 Getty Images.)

“We are trying to build an intelligent information machine that can achieve some of the computations that the human brain does, such as delicate motion controls, attention, reasoning, association, and decision making,” said Wei Yi, a principal investigator at HRL.

“For example, humans are much slower than current computers at arithmetic operations, in which all values have to be precisely calculated. But humans are superior to classical digital computers at intelligent tasks, such as fusing sensory data, mining data, filtering attention choices, induction and deduction, and even developing new concepts,” Yi said. “Human brains do not run on preprogrammed and fixed algorithms. The brain’s intricate, highly parallel network of 100 billion neurons and 100 trillion synapses is dynamic and adapts as it gains experience. At the end of the day, we are thinking about the possibility of building an adaptive electronic brain that might be able to self-learn. This seems a grand dream, but building biologically plausible neurons is a key to eventual memristive computers that mimic the cerebral cortex.”

MRAM deal
Spin Memory, an MRAM developer formerly known as Spin Transfer Technologies, has announced separate agreements with Applied Materials and Arm in the field of MRAM technology.

Spin Memory has developed various structures and perpendicular magnetic tunnel junction (MJT) technologies for use in STT-MRAM, a next-generation MRAM.

In the deal with Applied, Spin Memory and the equipment giant will collaborate and create an embedded MRAM solution. The goal is to develop an embedded MRAM manufacturing module, enabling the production of STT-MRAM products for use in embedded memory and SRAM-replacement applications. Spin Memory intends to make the solution available in 2019.

At the same time, Spin Memory has also announced an agreement with Arm under which Arm will license Spin Memory’s Endurance Engine design IP to address SRAM applications in chips. The engine is a combination of circuits and design architectures that are implemented in digital circuitry.

Spin Memory has also announced a $52 million Series B funding round. This funding round was led by Applied Ventures, the venture capital arm of Applied Materials, and Arm.