NANO: Information-Theoretic Foundation of Molecular Computing: Performance Limits and Design Optimization

分子计算试图通过探索纳米级组件之间的信息处理机制来推进新的计算范式。最终的系统将是超高密度、超快速的平台，能够执行可能带来直觉、创造力、效率和生物有机体其他基本特性的计算任务。这些未来的计算系统预计将通过使用生物编码遗传材料(DNA计算)或纳米级设备(如量子点设备、碳纳米管和纳米线(纳米电子计算))的自下而上自组装来构建。与传统的计算系统不同，分子计算具有一些独特的特点，如随机和不完全的分子相互作用，固有的冗余促进了大规模并行，以及分布式阵列结构促进了大量分子处理核心之间的合作。这项研究试图建立一个信息论基础，以揭示分子计算中基本挑战的关键性质和实际意义。这个项目的目标是：(1)建立分子计算的信息论模型；(2)使用信息论分析确定分子计算的基本性能极限；(3)探索新的分子计算解决方案，允许动态分配冗余的性能驱动任务和可靠性驱动的任务，从而将性能推向基本极限；以及(4)开发纳米电路和纳米级集成方案，将信息论的见解和物理设计技术结合到信息处理的新前沿。综合教育部分的目标是(1)满足计算机工程劳动力培训的更广泛需求，特别是新兴的分子计算技术；(2)培养研究生和本科生广泛参与探索分子计算的新想法；以及(3)在研究界、少数群体和普通公众中传播新的研究成果。

Molecular computing seeks to advance new computing paradigms by exploring information processing mechanisms among nanoscale components. The ultimate systems will be the platforms that are ultra-dense, ultra-fast, and capable of performing computational tasks that may bring intuition, creativity, efficiency, and other essential properties of living organisms. These future computing systems are expected to be constructed via bottom-up self-assembly using biologically encoded genetic materials (DNA computing) or nanoscale devices such as quantum-dot devices, carbon nanotubes and nanowires (nanoelectronic computing). Unlike conventional computing systems, molecular computing has some unique features such as random and imperfect molecular interactions, inherent redundancy facilitating massive parallelism, and distributed array architecture promoting cooperation among vast molecular processing cores. This research seeks to develop an information-theoretic foundation to reveal the key nature and practical implications of the fundamental challenges in molecular computing. The goals of this project are to (1) develop an information-theoretic model for molecular computing; (2) determine the fundamental performance limits of molecular computing using information-theoretic analysis; (3) explore new molecular computing solutions that allow dynamic distribution of redundant performance-driven tasks and reliability-driven tasks, thereby pushing the performance towards the fundamental limits; and (4) develop nanocircuits and nanoscale integration schemes that incorporate information-theoretic insights and physical design techniques for new frontiers of information processing. The goals of the integrated education component are to (1) address the broader need of workforce training in computer engineering, in particular, the emerging technology of molecular computing; (2) foster extensive involvement of graduate and undergraduate students in the exploration of new ideas of molecular computing; and (3) disseminate new research findings among research communities, minority groups, and general public.