COMPUTING WITH A TRILLION CRUMMY COMPONENTS.

This article explores the objective of building defect and fault-tolerant circuits on the nanometer scale. The author compares the built-in downtime and multiple runs needed to test data in the early days of computing and how John von Neumann and Claude Shannon introduced the concept of machines tha...

Descripción completa

Detalles Bibliográficos
Publicado en:Communications of the ACM Vol. 50; no. 9; pp. 35 - 40
Autores principales: Robinett, Warren, Snider, Gregory S., Kuekes, Philip J., Williams, R. Stanley
Formato: Artículo
Publicado: Association for Computing Machinery Sep2007
Materias:
Acceso en línea:Ver este registro en EBSCOhost
Descripción
Sumario:This article explores the objective of building defect and fault-tolerant circuits on the nanometer scale. The author compares the built-in downtime and multiple runs needed to test data in the early days of computing and how John von Neumann and Claude Shannon introduced the concept of machines that would work perfectly despite defective or unreliable components. Integrated circuitry achieved almost flawless precision, but miniaturizing circuits has brought the question of reliability back around to the components. How nanoscale parameters can complement CMOS development is discussed referencing Moore's Law rates of improvement while using only modest improvements in transistors.