The maximum CPU temperature rise with respect to the
initial temperature was approximately 3.0"C with 128 MB
RAM (fan-on). Increasing the RAM to 384 MB resulted in a
maximum temperature rise of approximately 6°C. With theCPU fan-off, the maximum temperature rise was about 10°C
using 128 MB in the RAM. This temperature rise jumped to
nearly 23OC when the RAM was increased to 384 MB.
As seen in Table 2, the temperature rises in the CPU are
normally low under fan-on condition; however, these could be
significantly high when the fan is off and the CPU has to work
intensively. Further, local temperature-rises within the CPU
could be even higher than the bulk temperature measured by a
single sensor such as the one used in our experiment.
Furthermore, as expected, the speed of program execution
increased with RAM amount as seen in Table 2. The least
processing speed (106 minutes) was achieved with RAM
amount of 384 MB. According to Figures 3-6 and Table 2, it
is possible to identify a RAM amount that provides optimal
values for both CPU temperature rise and speed of
calculation. For example, in our case, the RAM amount of
192 MB provides the optimal solution for both fan-on and
fan-off cases (see Table 2 and Figure 4).
In addition, it is clear from Table 2 that temperature rises
in hard disk are not significant for both scenarios. However,
it should be pointed out that for RAM amount,of 128 MB, the
hard disk was being constantly used as virtual memory which
is an U 0 intensive task. This in turn could lead to high
temperature and failure in hard disk for very prolonged
operations. It is also evident from Table 2 that RAM
temperature rises could be appreciable especially in the case
of fan-off.
The maximum CPU temperature rise with respect to theinitial temperature was approximately 3.0"C with 128 MBRAM (fan-on). Increasing the RAM to 384 MB resulted in amaximum temperature rise of approximately 6°C. With theCPU fan-off, the maximum temperature rise was about 10°Cusing 128 MB in the RAM. This temperature rise jumped tonearly 23OC when the RAM was increased to 384 MB.As seen in Table 2, the temperature rises in the CPU arenormally low under fan-on condition; however, these could besignificantly high when the fan is off and the CPU has to workintensively. Further, local temperature-rises within the CPUcould be even higher than the bulk temperature measured by asingle sensor such as the one used in our experiment.Furthermore, as expected, the speed of program executionincreased with RAM amount as seen in Table 2. The leastprocessing speed (106 minutes) was achieved with RAMamount of 384 MB. According to Figures 3-6 and Table 2, itis possible to identify a RAM amount that provides optimalvalues for both CPU temperature rise and speed ofcalculation. For example, in our case, the RAM amount of192 MB provides the optimal solution for both fan-on andfan-off cases (see Table 2 and Figure 4).In addition, it is clear from Table 2 that temperature risesin hard disk are not significant for both scenarios. However,it should be pointed out that for RAM amount,of 128 MB, thehard disk was being constantly used as virtual memory whichis an U 0 intensive task. This in turn could lead to hightemperature and failure in hard disk for very prolongedoperations. It is also evident from Table 2 that RAMtemperature rises could be appreciable especially in the caseof fan-off.
การแปล กรุณารอสักครู่..
