Nanoscale processors and interconnects
Intel was one of the first companies to see the benefits of nanoscale microprocessors, and will release their first mass-produced 45nm chips this year in full production. In the 1960s, the first Intel chips had just 30 transistors.
Up until the 1990s, Intel would scale the micro processor into smaller transistors for more power, but eventually found that additional scaling was not possible. Today, the microprocessor in your computer has millions of transistors contained on 20-24 chip layers.
The innovation involves nanoscale transistors - starting at 90nm, then 65nm, and now 45nm. In 2009, the company will move to 32nm, and then 22nm in 2011.
What's comes after that is not clear, and even at 45nm, there is still a leakage problem that engineers must offset in order to make the processor powerful enough, yet not use excessive energy. (This problem explains why both AMD and Intel have moved to multi-core chips.)
To understand how nanoscale processors work, it's important to explain briefly how a CPU is built. Chip designers make complex instructions for a CPU, specifying exactly where the nanomaterial must go on each wafer.
This specification is made into a mask; the mask has both transparent and opaque regions that expose or block light from reaching the silicon, thus forming patterns for the transistors. These patterns of silicon oxide are imprinted onto the wafer using a process called lithography.
The 'feature sizes' (or design) of the chip can be incredibly small: you can fit 2 million nano-transistors in the space of this period. The resulting wafers tend to warp slightly, so it is an exacting science.
Intel plans to produce transistors that are 30 per cent smaller every two years. With sub-40-nanometre transistors, there's a new issue with leakage because there is not enough insulation between transistors. Smaller transistors use only about 1 volt of power, compared to 5 volts used for older processors.
'The best way to think of it is a car engine,' says Rob Willoner, a technology analyst at Intel in the Technology and Manufacturing Group. "The more power you send to the engine, the more energy it consumes, and the more noise it generates. With processors, the challenge is to balance the power and energy use efficiently."
What's interesting about the new 45nm processors is the metal material used to make them - an ingredient that Intel has not specific yet, but will be forced to reveal it once the chips start being produced and delivered to consumers.
The concept of new materials resulting from nanoscale processors is one that could lead to some startling innovations, both with the processor and in the interconnects between the processor and the rest of the computer.
Marin Soljacic, an assistant professor of physics at MIT, says photonics - which involves nanoscale optical interconnects - could become the dominant method of connecting the CPU to the bus, RAM, and other components.
These nanoscale interconnects use a similar technology to the fibre optics used to deliver broadband Internet to homes. At the photonic scale, the interconnects could be much shorter than copper is now. This could lead to faster computers and less battery drain when moving the data inside the PC.
'There's a new material that develops when nanostructuring - these are optical materials that are not seen in nature,' says Soljacic. 'Photonics has other advantages as well when used with multiple parallel processors and multiple interconnects.'
Of course, while this all sounds a bit too much like The Matrix in that the processing would be 100 times faster than current generation technology and yet microscopic in size (and therefore mildly frightening!), the one major hurdle is fabrication (see sidebar). Processor plants would have to be retrofitted for new interconnect manufacturing, an endeavour that could take decades or may not become viable at all.
Exploding batteries, product recalls, firmware updates - these are just some of the reasons computer companies are looking at ways to provide long-lasting power for laptops and other devices, and yet keep the heat generation as low as possible.
Interestingly, the technology that could eventually replace the Lithium Ion battery is just a new, nano-scale version of the same concept. Li Ion works by arranging micron-size particles onto thin films. Companies such as A123 Systems have shown that these same particles can be arranged more efficiently at sub-100-nanometre sizes so that the charge lasts longer and does not generate as much heat.
'Think of Lithium Ion technology as a collection of marbles in a beach ball,' says Joe Adiletta, a product manager at A123. 'With smaller particles, they can move to the centre of the ball faster, because they do not have to travel as far. Travelling less distance means more speed, and more power, out of the same Lithium Ion process.'
Researchers at the Université Paul Sabatier have demonstrated how the nanoscale Lithium Ion battery technology could work for laptops and other mobile devices. The particles are arranged onto thin nanorod film material at a much higher density that micron-sized particles, which typically require a thicker film.
The result is a laptop or PDA battery that charges faster, lasts longer, but does not cause as much power drain as micron-size Lithium Ion. Like any new effort to find a solution for laptop mobility, nanoscale batteries will first have to prove their worth for consumer devices that generally last several days on one charge, but do not consume more than just a few watts of power.
Nano Emissive Displays
High-definition LCD and plasma displays are becoming more and more common, both in the living room and in the computer room. Yet, the colour quality on an old-style CRT display is still technically superior, even though these energy-draining monsters are heavy and cause excessive environmental damage.
Motorola has invented an LCD competitor called Nano Emissive Display which uses a cathode-ray tube for every pixel (instead of just one CRT gun on older monitors). The main benefit is in colour accuracy, since the technology can more accurately represent subtle dark shades than LCD (liquid crystal display) or DLP (digital light processing).
'With a Nano Emissive Display, each carbon nanotube pulls the electron through a field until it hits the glass surface,' says Jim Jaskie, the Motorola Chief Scientist. 'Because of this, they do not have the problems of LCD technology in motion and contrast, and the nanotubes generate very deep red and very deep blacks.'
With each carbon nanotube working independently, the display is more accurate in how it depicts movement, most noticeable in a hockey game where the puck is often lost in a sea of pixels on current displays, but would be much easier to trace on NED. A main challenge with NED is that it will require a new fabrication process.
Current LCD manufacturers have started using much larger motherglass sizes - which means a larger LCD or more inexpensive smaller displays cut from the original sheet. With NED, these plants would have to use a new nanotube fabrication process. Still, NED will be an attractive option because the manufacturing process costs less, the displays last longer than LCD, and the quality is noticeably superior.
Wireless technology - as we know it today - is all about fast throughput and wide coverage for home networking, and public hotspots for mobile computer access. In nanotechnology, the concept has more to do with microscopic power transmission.
The idea is to provide a way to charge a pocket device such as a cell phone just by placing it onto a charging pad and not connecting any wires.
Yet, a more futuristic use would involve charging just about any electric device from either close proximity - a base station that sends out a charge over a wireless connection to a receiver on the device - or even an entire city that is charged by a satellite broadcaster.
Like Wi-Fi technology, wireless power would use specific frequencies. With Wi-Fi, those frequencies (in the 2.4GHz or 5GHz spectrum) send data.
With wireless power, the frequencies would carry energy using a concept called resonance, which uses a focused beam of electromagnetic radiation. Your cell phone, PDA, laptop, hybrid car, or office building would receive these energy transmissions from a transmitter, similar to how a router sends a Wi-Fi signal to a laptop PC Card.
There's another way to think about wireless power transmissions - the nanoscale electromagnetic pulses for charging a device. RFID tags also perform a data transmission, but the receiving tag does not have it own power source, the transmitter also sends a "wake-up" signal to the tag, powers it during transmission, and then cuts both signals.
RFID tags are used for building security such that an employee can walk past a security checkpoint, have the card send identification information, and then pass by without ever having to swipe the card.
With wireless power, more information could be exchanged between the reader and the card, such as notifying the employee of important e-mails, adjusting a schedule for the time he or she enters the building, or even transmitting documents or rich media on the card.