5. Why Does The P4-M Permit A Higher Maximum Junction Termperature?

The observant reader will have noted that the values given for the maximum junction temperature of the mobile Pentium 4-M can be up to 33°C higher than those of a desktop CPU. Intel defines the maximum junction temperature as the hottest point on the processor surface, which determines the maximum temperature at which stable operation can be guaranteed. What at first glance appears to be a contradiction ("The mobile CPU has a lower TDP, but its junction temperature can be higher"), makes sense if we recall the following formula (it describes heat flux through a cooler in very simplified terms):

Iw = G*(TD-TA) with G=λ* (A/l)

where

Iw: heat flux (denotes any heat dissipation from the processor)

G: heat conductance
λ: thermal conductivity of the cooling material
A: area through which heat flows (contact surface between the die and cooler)
l: distance traversed by the heat flux
TD: maximum permissible die temperature according to specifications
TA: temperature around the ventilator

As can be deduced from the equation, the higher the temperature difference TD-TA is, the easier it becomes to transport a given heat flux Iw from the cooler and, thus, raise the maximum permissible junction temperature.

In other words: because the maximum junction temperature of a mobile CPU can be higher, the cooler has an easier time of it and does not require such an extravagant and space-consuming construction.

Advantage: The housing of a P4-M notebook can be designed to be slimmer and the devices flatter than those of a P4 desktop CPU.

Serious Disadvantage: Under certain circumstances, the processor surface can be substantially warmer, which is why a notebook equipped with it makes an excellent "thigh-warmer."

Pentium 4-M: Processors Optimized For The Job

As we've seen, a P4 desktop CPU is no different in design from its mobile counterpart. The distinction begins with the manufacturing process after the wafer production process, in other words, in the testing and validation, and the packaging, too. Energy-saving technologies such as Speedstep, Deep Sleep, etc., are also integrated into a P4 desktop CPU - however, these are not all activated, tested or validated. In addition, platform support is required for these technologies, such as BIOS support and the mobile variant of the I/O controller (e.g., ICH3-M) for Speedstep.

So what's behind the significantly higher price of a notebook CPU compared to its desktop equivalent?

Although the wafer production process is no different for either product, they go their separate ways in production once the dies are selected: the Pentium 4 is optimized for the highest possible computing power and clock speed, the Pentium 4-M for the lowest energy consumption. In the end, two fundamentally different products come off the line that differ in validation, housing and, naturally, their technical specifications.

Intel gives the following reasons for the relatively high price of its notebook CPUs:

The development of technologies such as Enhanced Speedstep, DeepSleep, DeeperSleep,
Quickstart, IMVP, etc., eats up a lot of money. These development costs of course trickle down to the individual processors. In addition, lot size of a mobile processor is far less than its desktop variant. For the comprehensive validation and selection (core frequency, supply voltage, Speedstep function, etc.), a great amount of additional work and process steps are necessary, which adds to the price.

We'll leave it up to the reader to decide whether these arguments hold water.