Original 45nm Intel® Core™ Microarchitecture

Greater Mobility Through Lower Power

INTRODUCTION

In the extremely diverse world of mobile platform design, the focus of Original Equipment Manufacturers (OEMs) is on “look & feel” (commonly referred to as Industrial Design). Mobile OEMs place great emphasis on creating unique system designs to differentiate themselves in the mobile market.[1,2] Differentiation results in systems that come in all shapes and sizes, or form-factors, as well as systems that emphasize different aspects of mobility such as size, weight, and features. This differentiation makes the specific industrial design of many notebook computers unique to their designer, and that uniqueness becomes associated with their brand. Therefore, components that make it easier for OEMs to implement unique designs allows them to achieve their brand goals.

Additionally, in order to build on the brand equity of a particular industrial design, OEMs frequently maintain the same mechanical design “skin” of the previous platform for two to three generations. This implementation choice also forces them to maintain the same thermal design characteristics as the previous design. Therefore, there is little room for an OEM to innovate when no other parameter is changed.

Although industrial design is a predominant OEM consideration, another major consideration is ergonomics. Ergonomics includes, but is not limited to, user touch-temperatures and audible system noise levels. An uncomfortable touch-temperature (also called chassis or “skin” temperature) or exhaust temperature, or an annoying system noise level distracts from the user experience, even for the sleekest systems. Given the highly integrated nature of notebook system designs, often the vectors of performance, noise, and comfort can be divergent, and each may constrain the other as OEMs seek to innovate and differentiate. OEMs spend substantial design effort to balance performance, chassis temperatures, and or quiet systems, as well as differentiating along one or more of these vectors.

Intel's introduction of lower-power, high-performance Intel® processors based on original 45nm Intel Core™ microarchitecture, originally referred to by the codename Penryn, to the mainstream mobile market brings a new level of opportunity for differentiation. The mainstream Penryn mobile processor draws 25 watts (W), 10 W less than its 35-W predecessor. In this paper we focus on those attributes of the design that are enhanced by lower power Penryn family mobile processors: 1) thickness, 2) temperature, and 3) noise level. We concentrate on the mainstream Penryn family of mobile processors and do not describe the benefits of utilizing the even lower power, small-form-factor family of Penryn mobile processors.

In the following sections, we show form-factor trends and summarize fundamental limits of cooling under ergonomic constraints and other boundary conditions. We then relate these limits to form factor, noise, system temperatures, opportunities for feature additions, and design flexibility.

Current Platform Thermal Challenges

Mobile OEMs design notebook cooling solutions with challenging form-factor limits, component temperature limits, and the primary ergonomic boundary conditions of noise level and chassis “skin” temperatures. A representative system layout is illustrated in (Figure 1).

Figure 1: Representative notebook iso-view with transparent chassis

The numerous subsystems and components compete with each other for motherboard and notebook perimeter real estate for connectors and user interfaces. The result is little unused space, leaving marginal room for adequate air flow. As a consequence, systems are highly integrated mechanically and thermally; they are interdependent for balancing the limited cooling available. The major cooling components, the heat pipe, fan, and heat exchanger, are often referred to as the thermal solution. However, in reality, the solution is the total system design itself, as careful consideration is given to the placement of the air inlet and exhaust vents, the placement of components in primary air flow paths, air flow paths themselves, the fan and heat exchanger, as well as conduction interfaces such as spreaders and even insulators. Change in any of these parameters impacts the cooling of components and the system as a whole, which can make or break a successful design.

Within this competitive landscape, the higher power components vie for good air flow paths, and in the best case, a direct attach to the active solution (the fan and heat sink assembly). Since these good air flow paths and the placement options of the active solution are limited, the higher the power of the component, the more restrictions it places on the system design, whether in the location of the device itself, or in limiting the ability to attend to other components.

Thermodynamic limits

In any given system there is a maximum sustainable level of cooling that is established by the thermodynamic limit. This limit includes theoretical passive limits and the capacity of the available air flow to absorb heat between the ambient air (input) temperature and any given maximum allowable exhaust (output) temperature.

Most of the powered components reside in the base of the system, and the base represents the primary cooling challenge in mobile systems. An energy balance about the base of the system in steady-state conditions is

where P base is the allowable total power of components in the base of the system; that is, excluding power to the display. Q passive is the combined passive heat transfer mechanism for energy dissipation; that is, radiation and natural convection, represented by a form of Fourier's Law.

h is a combined effective heat transfer coefficient including natural convection and an approximation of radiation; a typical value may be 8 W m ² -K. Considering Equation 2 , once the system form factor, and thus A (area), is fixed, and once the ambient temperature and the allowable maximum skin cooling temperature (an ergonomic limit) are defined, the passive cooling is established.

Q active is the active cooling level, and its maximum capability is determined by the thermal capacity of the air flowing through the system as in

Equation 3 illustrates how, in a given ambient temperature, and a maximum allowable exhaust air temperature, the maximum active cooling capability is defined wholly by the amount of air that can be passed through the system. The relation is linear. (Figure 2) shows this relationship with the thermodynamic limit for a 1.1”-thick, 14”-display system with nominal boundary conditions of 35°C ambient, an average chassis temperature of 50°C, and assuming a maximum exhaust air temperature of 70°C.

Figure 2: Maximum sustainable system cooling—thermodynamic limit

Air flow then is the most critical determinant of the total cooling capability of the system; it carries the primary burden for cooling the system. How much air flow can be passed through the system, of course, relies more on the performance of the fan. This capability is primarily characterized by the allowable space for the fan and air flow paths around the fan. There is limited area on the motherboard as well as limited available chassis perimeter space for the fan exhaust, so the only remaining major parameter to examine is the available internal height for the fan and air flow around it.

Figure 3: Side-view cutaway of vertical stack-up (base only) for a nominal 1.1” system

Thermal stack-up

In mobile systems, the amount of internal height available for the fan and air flow approaching it is limited to, and roughly characterized by, the vertical distance z-height between the bottom of the keyboard and the inside bottom chassis surface. This space is a fraction of the total stack-up as shown in (Figure 3) . For example, in a typical 14”-display and a 1.1”-thick system, the internal z-height available is 15mm, which is broken down roughly to 12mm for the fan itself and 3mm for air flow paths about the inlet(s). For a system of this size, a typical maximum air flow rate may be 3.3 cfm (1.6e-3 m ³ s); of course, actual air flow depends on the specific design, particularly on system flow paths and corresponding flow resistances.

Several of the system stack-up components at any given time are virtually fixed, such as the display, the keyboard, the motherboard, and the chassis skin. Consequently, if an approximate 10-percent or 0.1” (2.5mm) reduction in total system thickness is desired, it is effectively a reduction in the vertical space available for the fan and air flow paths at the inlet, or approximately 20 percent for the fan itself (2.5mm out of 12mm) assuming the space for the air flow inlet is retained.

Thermal technology limits

Cooling technologies are focused on increasing efficiency in component cooling interfaces and on air flow, whether it's the fan itself or some other means. These technologies have resulted in roughly only a 5-percent additional total system cooling capability year-over-year. Moreover, this progression is prone to tapering off over time with diminishing returns on incremental improvements. Therefore, cooling technology alone cannot keep up with a desire for progressively thinner systems. The power of the components themselves must therefore be reduced to realize thinner systems.

Back to Top