Page 16

EETE NOV 2015

nanotube transistors Schematic showing a fabricated set of devices with different contact geometries on the same nanotube to verify that the contact size can shrink without reducing device performance. (Source: IBM Research, used with permission) nique and thinks it will give IBM a competitive advantage over the competition. It works by coating the end of the nanotubes with molybdenum before their self-aligned placement as transistor channels. Then the ‘welding’ step is completed by heating the whole assembly to 1,562 degrees Fahrenheit (850 degrees Celsius) thus melting the molybdenum and chemically transforming it into carbide, a conductor that makes an excellent angstrom-scale contacts. “This clever welding/carbide discovery seems to be the only silicon-processing favorable metal without oxidation problems and or Schottky barrier characteristics would add an undesirable step-voltage threshold. Rather IBM’s method has an on/ off current switching ratio of 10,000, which is great,” added Doherty. “Others trying side connection schemes to nanotubes seem to vary in resistance and conductance over time, making them only lab curiosities. IBM’s end-cap success paves the way for real nanotube structures that can make use of already proven silicon photolithography techniques. Today IBM is only able to produce p-type transistors using its revolutionary “welding” technique, but it claims the devices work better than FinFETs and their next step--already under way with light at the end of the tunnel, according to Han--is n-type devises so that CMOS nanotube transistors can be fabricated at the 3-nanometer node by the time silicon FinFETs reach the 5 nanometer node. “IBM’s internal goal is to be ready to come in at the 5-nanometer node and become the best option for the 3-nanometer node all the way down to the angstrom level,” Hans told us. According to Doherty, IBM’s achievement gives new hope to extending Moore’s Law into the indefinite future, since no insurmountable scaling problems are seen for the nanotube ‘welding’ technique, which Doherty dubs IBM’s “coaxial connection” technology. “And succeeding with these coaxial end caps at nine nanometers gives courage and confidence to the industry to there being an alternative semiconductor path as silicon gets more and more inefficient as we push its limits under Moore’s Law,” Doherty said. “Also I have rarely seen a paper IBM’s Science paper with so many attributions to other researchers. Their goal is production within a decade, but I am sure they want it faster than that if possible.” IBM’s success is a part of its $3 billion research and development effort for post-silicon technologies, started in 2014, to pave the way to the computing needs of what IBM calls the coming “cognitive computing era.” Cyber threats against cars are here to stay, experts say By Christoph Hammerschmidt The rise of the connected car definitively catapults our traditional set of wheels into the world of information technology. This arrival confronts carmakers with an unfamiliar challenge: The possibility of getting hacked. At a meeting in Dresden (Germany), experts analysed the threat and potential approaches to harden the vehicles against malicious attacks. Losing control over your vehicle is a very frightening experience. The video of a driver helplessly cranking the wheel while his Jeep Cherokee drove into the ditch, remotely controlled by a hacker duo, scared the world. This orchestration, in the headlines as recently as past August, should have been the last wake-up call for automotive electronics designers: Car can get hacked, and they should move quickly to lock any hackers out. At the Cyber Secure Car industry meeting this week in Dresden, experts from the automotive value chain, the academic research community and the electronics and software industry discussed the current position of the car industry in terms of cyber security, hacker motivations and practices and potential measures to make cars more secure and - since in the car security is equivalent to safety - safer. Carmakers and their suppliers urgently need to make themselves knowledgeable about the risk, this was the unanimous position of the presenters. Given the average age of the cars out on the streets of some ten years, currently only a small percentage of this “installed bas” already possesses a wireless interface to the outside world. But the number of connected cars will increase steeply over the next decade, and malicious attacks will increase in parallel, predicted Florent Frederix from the Online Trust and Cyber Security unit of the European Commission’s Directorate-General for Communications Networks, Content and Technology. But there is no need to wait for the connected car, already all contemporaneous vehicles have some kind of wireless interface that can potentially used for an attack. “There are more than 50 attack points in the connected car ecosystem”, said Mike Parris who oversees the Secure Car Division at British telematics and automotive security consultancy SBD. “Any point of the ecosystem can be hacked”. Like Frederix, he highlighted the parallelism of the connected car with the Internet of Things. “After Stuxnet (the computer virus that in- 16 Electronic Engineering Times Europe November 2015 www.electronics-eetimes.com


EETE NOV 2015
To see the actual publication please follow the link above