2-D Crystal Electronics for Energy-Efficient Terabit Communication Links

到2007年，在iPhone和iPad推出之前，美国ICT行业的碳足迹已经超过了航空业。从那时起，无线通信设备、社交媒体(Facebook、Twitter、LinkedIn等)的激增。而云计算只会加剧ICT行业能源消耗的年增长率。我们现在在全球拥有2049个数据中心，消耗300亿瓦，相当于30个核电站。按人均计算，加拿大人在互联网使用率和社交媒体网站注册人数方面领先于美国人。我们对无线通信的胃口似乎永无止境。不幸的是，在所有的通信链路中，无线是效率最低的。对于计算来说，前景同样黯淡。在过去的12年里，缺乏晶体管亚阈值斜率的改善，以及缺乏适当的电源电压调节，这意味着微处理器的功率密度已经变得有限。虽然并行和多核微处理器体系结构暂时缓解了这一问题，但代价是芯片上的通信增加和时钟频率停滞，只有发明具有更陡峭的亚阈值斜率、更快的传输和更高的电流驱动能力的新型晶体管，才能在0.3至0.1 V的电源范围内运行，才能遏制或扭转功率密度的趋势。 这项研究计划的重点是开发(I)用于太比特数据速率短距离无线通信的新型亚毫米波(250-500 GHz)高效数字发射机以及(Ii)新型纳米级金属纳米线和2-D晶体晶体管，旨在将数据中心和无线链路的能效提高2-3个数量级。研究成果将包括10 nm以下晶体管的实际实施和运行速度接近0.5Tb/S的短距离无线链路。这些革命性的设备和无线链路将有助于重新确立加拿大在ICT领域的领先地位，并将带来全新的技术、产品和公司。

By 2007, before the iPhone and the iPad were introduced, the US ICT industry already had a larger carbon footprint than the airline industry. Since then, the proliferation of wireless communication devices, of social media (Faceboook, Twitter, Linkedin, etc.) and cloud computing has only exacerbated the annual growth rate of energy consumption attributed to the ICT industry. We now have 2049 data centers worldwide which consume 30 billion Watts, the equivalent of 30 nuclear power stations. On a per capita basis, Canadians rank ahead of Americans in terms of Internet usage and enrollment in social media websites. Our appetite for wireless communication seems insatiable. Unfortunately, of all communication links, wireless is the least efficient. The outlook is just as bleak for computation. Lack of transistor subthreshold slope improvement and lack of appropriate supply voltage scaling during the last 12 years has meant that microprocessors have become power-density limited. While parallel and multicore microprocessor architectures have temporarily alleviated the problem at the expense of increased communication across the chip and stalled clock frequency, only by inventing new transistors with steeper subthreshold slopes, faster transport, and higher current driving capability, which will permit operation from 0.3 to 0.1 V supplies, will the power density trend be curbed or reversed. With a focus on the development of (i) new submillimetre-wave (250-500 GHz) high efficiency digital transmitters for terabit data rate short range wireless communication and of (ii) novel nanoscale metal nanowire and 2-D crystal transistors, this research proposal aims to increase the energy efficiency of data centres and of wireless links by 2-3 orders of magnitude. The research output will consist of the physical implementation of sub-10nm transistors and of short range wireless links operating at close to 0.5 Tb/s. These revolutionary devices and wireless links will help to re-establish Canada's position as a leading player in the ICT field and will lead to radically new technologies, products and companies.