Desktop Microprocessors Research Articles

Functionality Versus “Typical Product” Measures of Technological Progress

Summary Technological progress and adoption are fundamentally interconnected with environmental challenges faced by society. At the product level, researchers often explore the interplay between technological change and the environment by tracking trends in impacts per unit functionality—for example, gasoline consumed per distance traveled by a vehicle. In this article, we explore an alternative measure: “typical product.” A typical product measure accounts for changes in consumers’ demand and use of products as product quality improves—for example, gasoline consumed for a typical driving pattern for a vehicle. We compare and contrast functionality and typical product measures through a case study of electricity use to fabricate Intel desktop microprocessors from 1995 to 2006. The functionality normalization is measured in terms of electricity use per transistor produced. Results show rapid and sustained exponential decrease. The “typical product” measures electricity use per typical desktop microprocessor of a given year (e.g., a Pentium II in 1998, a Pentium IV in 2002). Results show that, despite fluctuations, energy use per typical microprocessor is roughly constant over the 12-year period. The explanation of this result is that although technological progress dramatically reduces the energy needed per transistor, it also induces demand for more powerful chips, which contain many more transistors. The typical product measure has applications in defining functional units in life cycle assessment, characterizing rebound effects, and measuring energy efficiency trends.

Heatsink Mass Optimization Methodology for Desktop Microprocessor Cooling

We present a novel modeling methodology for optimizing the mass of aluminum-extruded heatsinks for cooling desk-top microprocessors. The two-stage study aims at reducing mass of an aluminum-extrusion heatsink by taking advantage of the existing thermal resistance margin. In Stage 1, we investigated several possible combinations of base shapes to minimize the mass of the base region. Base shape is optimized by dividing the base volume into several blocks of equal width, with the objective of progressively minimizing the mass of the peripheral region while shaping the middle region to minimize spreading resistance. Three base shape profiles—namely, wedge, semielliptical, and flat top—were selected for further study. In Stage 2, a thorough base, fin thickness, and fin count study was carried out by mathematical optimization. The result of this optimization exercise was a 30% reduction in the heatsink mass. The degradation in thermal resistance of the mass-reduced design is within the acceptable margin and meets microprocessor product specifications. Results are experimentally verified by testing four different competing samples having comparable optimized masses.

A Numerical Study of the Thermal Performance of an Impingement Heat Sink—Fin Shape Optimization

This paper presents the results of a numerical analysis of the performance of an impingement heat sink designed for use with a specific blower as a single unit. These self-contained heat sink/blower units, which cause impingement type flow on the heat sink fins, are now commonly used for desktop microprocessors. One of the objectives of this study is to examine the effect of the shape of the heat sink fins, particularly near the center of the heat sink, on the thermal performance of the package. The pressure gradient at the center of the heat sink, near the base, tends to be high. It significantly reduces the airflow, and hence, transport in that region. Different fin shapes and airflow rates have been studied with the objective of searching for an optimal heat sink design that would improve the thermal performance without increasing the pressure drop across the heat sink. Parallel plate fins have been studied by removing fin material from the region near the center of the heat sink along the length and height of the fins. Seventeen different designs have been compared, and an "optimum" heat sink shape is reported that results in a lower operating temperature and pressure drop. It is found that removal of fin material from the central region of the heat sink enhances the thermal as well as hydraulic performance of the heat sink.

Designing and verifying embedded microprocessors

Motorola's ColdFire products are a line of microprocessors targeting embedded-system applications such as computer peripherals (disk drives, laser printers, scanners).They originated from the same design group that produced the 680X0 general purpose microprocessors, whose target market was desktop computing applications. The ColdFire microprocessors, however, target the highly competitive embedded market, whose time-to-market and cost requirements are much more stringent. The ColdFire design team received a set of challenges quite different from those associated with the 680X0 line. They had to minimize test costs since the target selling price was an order of magnitude less than that of a desktop microprocessor. With no room for design errors, they had to put processes in place that detect errors early in the design flow and provide feedback for continuous improvement. A new methodology reduced new product cycle time to less than a year for the ColdFire products and provided improved techniques for integrating cores in new applications. In addition, it increased quality measurement capability and reduced test cost.

Interfacing a modern frequency response analyser with a desk-top microprocessor

The practice of using a microprocessor to facilitate data logging and experimental control is well established. However, there are frequently difficulties in establishing communications between a device and a computer. An approach to interfacing the Solartron 1250 frequency response analyser and the Kemitron K2000E microprocessor is described.