Touch and go for Moore? Moore's law - the observation that the number of transistors in a dense integrated circuit doubles approximately every 18-24 months - is a projection and not a physical or natural law. Andy Pye looks at some of the disruptive technologies which aim to ensure it can continue.
But can this continue indefinitely? Intel stated in 2015 that the pace of advancement has slowed, starting at the 22nm feature width around 2012, and continuing at 14nm. However, in April 2016, Intel CEO Brian Krzanich stated: "In my 34 years in the semiconductor industry, I have witnessed the advertised death of Moore's Law no less than four times. As we progress from 14nm technology to 10nm and plan for 7nm and 5nm and even beyond, our plans are proof that Moore's Law is alive and well."
Silicon-based technologies have nearly reached the physical limits of the number and size of transistors that can be crammed into one chip, while alternative technologies are still far from mass implementation. Down-scaling transistor size is more than an engineering challenge, as there is fundamental physics to consider.
So overall, perpetuating Moore's Law in the foreseeable future will require disruptive technologies which take the electronics industry beyond its silicon comfort zone. Whether these can be developed in time for Moore's Law to be maintained in the immediate future is touch and go, but even if not, there is no reason why it could not resume in the future.
"The whole semiconductor industry wants to keep Moore's Law going. We need better performing transistors as we continue down-scaling, and transistors based on silicon won't give us improvements anymore," says Heinz Schmid, a researcher with IBM Research GmbH at Zurich Research Laboratory in Switzerland.
Schmid's team with support from colleagues in Yorktown Heights, New York has developed a relatively simple, robust and versatile process for growing crystals made from compound semiconductor materials that will allow them to be integrated onto silicon wafers - an important step toward making future computer chips that will allow integrated circuits to continue shrinking in size and cost even as they increase in performance.
The IBM team has fabricated single crystal nanostructures, such as nanowires, nanostructures containing constrictions, and cross junctions, as well as 3-D stacked nanowires, made with so-called III-V materials. Made from alloys of indium, gallium and arsenide, III-V semiconductors are seen as a possible future material for computer chips, but only if they can be successfully integrated onto silicon. So far efforts at integration have not been very successful.
"What sets this work apart from other methods is that the compound semiconductor does not contain detrimental defects, and that the process is fully compatible with current chip fabrication technology," says Schmid. "Importantly the method is also economically viable."
ZERO RESISTANCE MATERIALS
As miniaturisation progresses, power becomes critically important: how to reduce power flowing through electronic components? Related to this is the amount of heat generated during operation.
While atomic and molecular sizes cannot be changed, the heat problem is not unsolvable. Recent research has shown that in two-dimensional systems, including semiconductors, electrical resistance decreases and can reach almost zero when they are subjected to magnetic and microwave influence.
There are several different models and explanations for the zero-resistance phenomenon in these systems. However, the scientific community has not reached an agreement on this matter because semiconductors used in electronics are complex and processes in them are difficult to model mathematically.
An example is research conducted by the Quantum Dynamics Unit at Okinawa Institute of Science and Technology Graduate University (OIST), which could represent an important step in understanding two-dimensional semiconductors. The Unit is researching anomalies in the behaviour of electrons in a liquid helium two-dimensional system.
The system is maintained at a temperature close to absolute zero (-272.75[degrees]C or 0.4K) to keep the helium liquefied. Conditions are similar to those that led to observations of zero resistance in semiconductors.
SPIN NOT CHARGE
One promising approach to developing new technologies is to exploit the electron's tiny magnetic moment, or 'spin'. Electrons have two properties-charge and spin - and although current technologies use charge, it is thought that spin-based technologies have the potential to outperform the charge-based technology of semiconductors for the storage and processing of information.
Scientists from University College London (UCL) have discovered a new method to efficiently generate and control currents based on the magnetic nature of electrons in semi-conducting materials, offering a radical way to develop a new generation of electronic devices.
In order to utilise electron spins for electronics, or "spintronics", the method of electrically generating and detecting spins needs to be efficient so the devices can process the spin information with low power consumption. One way to achieve this is by the spin-Hall effect, which is being researched by scientists who are keen to understand the mechanisms of the effect, but also which materials optimise its efficiency. If research into this effect is successful, it will open the door to new technologies.
Meanwhile, researchers at Cambridge University have built a miniature electro-optical switch which can change the spin - or angular momentum - of a liquid form of light by applying electric fields to a semiconductor device a millionth of a metre in size. Their results demonstrate a way to bridge the gap between light and electricity, which could enable the development of ever faster and smaller electronics.
There is a fundamental disparity between the way in which information is processed and transmitted by current technologies. To process information, electrical charges are moved around on semiconductor chips; and to transmit it, light flashes are sent down optical fibres. Current methods of converting between electrical and optical signals are both inefficient and slow, and researchers have been searching for ways to incorporate the two.
University of Cambridge researchers, led by Professor Jeremy Baumberg from the NanoPhotonics Centre, in collaboration with researchers from Mexico and Greece, have built a switch which utilises a new state of matter called a polariton Bose-Einstein condensate in order to mix electrical and optical signals, while using miniscule amounts of energy.
Polariton Bose-Einstein condensates are generated by trapping light between mirrors spaced only a few millionths of a metre apart, and letting it interact with thin slabs of semiconductor material, creating a half-light, half-matter mixture known as a polariton.
Putting lots of polaritons in the same space can induce condensation - similar to the condensation of water droplets at high humidity - and the formation of a light-matter fluid which spins clockwise (spin-up) or anticlockwise (spin-down). By applying an electrical field to this system, the researchers were able to control the spin of the condensate and switch it between up and down states. The polariton fluid emits light with clockwise or anticlockwise spin, which can be sent through optical fibres for communication, converting electrical to optical signals.
While the prototype device works at cryogenic temperatures, the researchers are developing other materials that can operate at room temperature, so that the device may be commercialised.
|Printer friendly Cite/link Email Feedback|
|Date:||Apr 1, 2017|
|Previous Article:||The final frontier! Space is emerging as a new frontier for private investors. Andy Pye looks at the move to take commercial aerospace beyond its...|
|Next Article:||More skin in the game: automating cleanrooms comes with a new set of challenges. How do these systems operate effectively without risk of...|