New Material Pave the Way for Super fast optical computing

New phase-change material lights the way to all-optical, super-fast computing

British researchers have created a new material that could allow for the creation of all-optical computers — computers that are orders of magnitude faster and power efficient than today’s hot, sweaty electronic beasts. This isn’t some crazy, hard-to-fabricate material like graphene, either — we’re talking about a special breed of chalcogenide glass.

You’d be forgiven for not knowing what a chalcogenide glass is, but trust me, you’ve all seen one — and probably made extensive use of them, too. Chalcogenide glasses are a family of (rather oddly behaved) materials that can readily change state between glassy and crystalline, which also changes their optical and and electrical properties. The recording layer of every rewritable CD and DVD consists of a chalcogenide glass (usually germanium-antimony-tellurium, GST) that changes state between glass and crystal when it’s struck by powerful laser, or can have its state read by a weaker laser, and can thus be used to store and retrieve binary data. Chalcogenide glasses are also used in modern, super-fast phase-change memory (pictured above), where a small heating element is used instead of a laser to change the state of a memory cell.

In modern computing, there’s a strong dichotomy between optical and electronic systems. On the one hand, almost every computer is governed by the flow of electricity — along copper wires and over transistors. The flow of electricity creates a lot of heat, though, which limits how fast a computer can operate — and separately, electrical resistance/impedance makes electricity poorly suited for long-range transmission (you need to pump the voltage up to go more than a few meters, and that causes issues). This is why we use lasers and optical fibers for backbone internet connections — it’s both faster and more efficient.

Read: The secret world of submarine cables

We would love to use lasers instead of electricity inside our computers, but a) it’s hard to build tiny lasers that fit on a small silicon chip, and b) optical transistors with comparable performance to the latest 14nm FinFET transistors are hard to come by. Now, researchers at the University of Surrey, University of Cambridge, and University of Southampton in England may have come up with a solution that will allow for all-optical computers — a single component that can produce light, guide light, store light, and detect light. [Research paper: doi:10.1038/ncomms6346 / pre-print PDF]

Basically, the British researchers have taken a chalcogenide glass — which are always p-type semiconductors — and managed to turn it into an n-type semiconductor. (In this case, they germanium telluride (GeTe) and ionically doped it with bismuth to create an n-type). This allowed them to build a p-n junction — the primary building block of almost every semiconductor device, from transistors to LEDs to photovoltaic cells.

The research paper seems to stop short of actually usingthese chalcogenide p-n junctions, merely marveling at the fact that they’ve managed to create them in the first place — but the accompanying press release is pretty optimistic about their world-changing applications. “The challenge is to find a single material that can effectively use and control light to carry information around a computer. Much like how the web uses light to deliver information, we want to use light to both deliver and process computer data,” says project leader Richard Curry. “In doing so, this could transform the computers of tomorrow, allowing them to effectively process information at much faster speeds.”

As for when we’ll actually see chalcogenide transistors and all-optical computers, we’re probably talking years. Phase-change memory, however, which provides non-volatile storage with performance that’s better than DRAM, is right on the cusp of commercial availability.

Optical Computing At a Speed of light

By 2020, you could have an exascale speed-of-light optical computer on your desktop

Optalysys, a UK technology company, says it’s on-target to demonstrate a novel optical computer, which performs calculations at the speed of light, in January 2015. If all goes to plan, Optalysys says its tech — which is really unlike anything you’ve ever heard of before — can put an exascale supercomputer on your desk by 2020.

When we talk about optical computing, we’re actually referring to a fairly large number of different and competing technologies. At its most basic, optical computing refers to computing that uses light instead of electricity. When we’ve previously written about optical computing, we’re usually referring to chips and computers that have replaced their internal wiring with optical waveguides, and some kind of optical transistor that is controlled by photons instead of electrons. There are also optoelectronic devices, which use a mix of the two (usually optical interconnects and electronic transistors).

In the case of Optalysys, optical computing is something else entirely. At this point, because the Optalysys tech is rather complex, you should probably watch the video embedded below — not only will it probably do a better job than me at explaining it, but it’s also narrated by the adorable Heinz Wolff. If you can’t watch the video, read on and I’ll try my best.

It goes something like this. You start with a low-power laser. This laser is then directed through a massive liquid crystal grid. This grid works in much the same way as a liquid crystal display. By applying electricity to each “pixel,” the laser light passing through it is affected. Complex calculations would turn hundreds or thousands of these pixels on or off. After the laser has passed through this grid, the beam is picked up by a receiver. By analyzing the beam’s diffraction and Fourier optics, matrix multiplication and Fourier transforms can be combined to perform complex maths. You can also have multiple pixel grids in sequence or parallel, significantly boosting the complexity and parallelism of the optical computer. There’s a little more technical info on the Optalysys website, but not much.

Moving away from the technical nitty-gritty, Optalysys’s optical computer is exciting for two main reasons: It consumes very little power, and there’s essentially no limit on how parallel you can make it. There’s no direct analogy to transistor-based logic, but you could almost think of every liquid-crystal pixel as a tiny processing core (or at least a tiny transistor). In a normal computer chip, while there is some parallelism, most things happen very sequentially, with each core (and each transistor) working mostly in serial. In an Optalysis optical computer, the laser beam hits every single pixel at the same time — it essentially performs hundreds or thousands (or millions?) of small computations in parallel, at the speed of light.

In terms of power consumption, HPCwire says the estimated running costs of an Optalysys computer would be in the region of $3,500 per year. I think this is based on the assertion by Optalysys that it will be able to deliver a desktop-size computer with exascale performance. Compare this to the world’s fastest supercomputer, which would cost around $21 million in energy costs if it was run at peak (~34 petaflops) performance for a year.  [Read: HP bets it all on The Machine, a new computer architecture based on memristors and silicon photonics.]

Optalysys says that its technology is already at NASA Technology Readiness Level (TRL) 4, meaning it’s ready for full-scale testing in a laboratory environment. By January 2015, Optalysys says it will have a 340-gigaflop prototype ready to go. By 2017, the company wants to have two commercial systems in place: A big data analysis system that will add 1.32 petaflops of grunt to an existing, conventional supercomputer — and a standalone Optical Solver supercomputer, which will start at 9 petaflops. Optalysys thinks its Optical Solver could scale up to 17.1 exaflops by 2020. This seems like a very bold statement for an entirely novel and untested method of computing — but given how conventional computing has mostly stagnated by this point, I hope the folks at Optalysys can follow through.

Cryogenic on-chip quantum electron cooling leads towards computers that consume 10x less pow

Cryogenic Technology Helps Improve Cooling And Reduce Power Consumption of Computers

Researchers at UT Arlington have created the first electronic device that can cool electrons to -228 degrees Celsius (-375F), without any kind of external cooling. The chip itself remains at room temperature, while aquantum well within the device cools the electrons down cryogenic temperatures. Why is this exciting? Because thermal excitation (heat) is by far the biggest problem when it comes to creating both high-performance and ultra-low-power computers. These cryogenic, quantum well-cooled electrons could allow for the creation of electronic devices that consume 10 times less energy than current devices, according to the researchers.

What, you may ask, is a quantum well? In essence, a quantum well is a very narrow gap between two semiconducting materials. Electrons are happily bouncing along the piece of semiconductor when they hit the gap (the well). Only electrons that have very specific characteristics can cross the boundary. In this case, only electrons with very low energy (i.e. cold electrons) are allowed to pass, while hot electrons are sent back from whence they came. (If you’re technically minded, the well is created by sandwiching a narrow-bandgap semiconductor between two semiconductors with a wider bandgap – it’s basically the quantum equivalent of the neck between the two bulbs of an hourglass).

Read: The fanless heatsink: Silent, dust-immune, and almost ready for prime time

Once you have a stream of cold electrons coming through the quantum well, you can tack other components onto the circuit — such as transistors. In this case, to prove that their quantum well was actually producing cold electrons, the UT Arlington researchers created some single-electron transistors (SETs) — very efficient transistors that are very sensitive to thermal excitation and thus usually have to be cryogenically cooled with liquid helium. Despite being at room temperature, the SETs worked just fine — all thanks to the 45 Kelvin (-228C) electrons. The chip pictured at the top of the story, created by the UTA researchers, contains a bunch of quantum wells and SETs.

That’s me, checking out one of IBM’s quantum computers

Moving forward, it now becomes a question of integrating these quantum wells with actual electronic devices. “When implemented in transistors, these research findings could potentially reduce energy consumption of electronic devices by more than 10 times compared to the present technology,” says Usha Varshney of the National Science Foundation, which funded UTA’s research. Because every facet of computing is interconnected by the laws of physics, a 10x reduction in power consumption could affect just about everything — from battery life, to device size and weight, to max performance. [Research paper: doi:10.1038/ncomms5745 – “Energy-filtered cold electron transport at room temperature”]

The researchers also note that further investigation into quantum well construction could result in a stream of electrons with zero temperature (i.e. zero kelvin). As we’ve discussed before, weird and wonderful things can happen when you get down towards absolute zero — not least of which, the construction of some wondrously powerful and energy efficient computers.