|
Página 11 de 14
Test equipment: - Motherboard: DFI LP NF4 Ultra D
- Processor: AMD 64 Venice 3000+
- Memory: 2 x 512 Corsair XMS Expert
- Power Supply: Sunbeamtech NUO 500W
This is a standard configuration for a modern PC, but what will not be standard are the work conditions in which the kit’s performance is going to be tested. We will go from 200 x 9 @ 1.4V to 300 x 9 @ 1.65V, ( http://valid.x86-secret.com/show_oc?id=96464 ) an excessive voltage for that frequency, it is true, but the aim here is to increase the processor’s temperature far beyond any normal condition.
All the tests where run in a constant room temperature of 24°C. To have an idea of the kit’s performance, we have compared it with AMD’s cooler stock as we think it to be the only cooler that is within the reach of any user, with no exception. In this way, you will be able to draw your own conclusions, since if we compared it with any other cooler, it would not be possible for the user to estimate the behavior applied to his own case, no matter how massive that particular cooler may be, and as a result, only some of them will know how it will really perform.
These are the temperatures we got with the cooler stock and making the processor
work in the following way:
with the above mentioned setting, the processor in idle state and a stable temperature, we ran the well-known Super Pi bench with 32 M. At the same time, all the other PC processes continued running (antivirus, Internet, email program, etc.) to ensure the processor went from ‘idle’ to a completely demanded state, and with a 100 % workload. It is important to point out that the temperatures were taken using the sensor included in the hardware and monitored with ITE, a specific program that comes with the motherboard.
As we can see, the result with the cooler stock is quite an alarming one, reaching 62°C in some moments. This is far from what we are accustomed to tolerating in our own PCs.
In this second chart, we can see the evolution of the temperature in the same conditions shown above, but this time with the Intro Kit installed. To be more graphic with the result comparison, we have left that of the cooler stock. The temperature decrease was remarkable: 20C difference under the same work conditions and with a full load processor. Not bad at all, if we take into account the temperature range we were working with. With the cooler stock we were dangerously close to the processor’s thermal tolerance limit while with the Kit, that range was totally acceptable, considering the high frequency and voltage given to the processor.
It is not convenient to analyze the kit from the overclock a processor achieved, since it is not enough to have a low temperature to reach a high overclock. There are many variables to be considered which do not depend on temperature.
However, if you wonder which the maximum stable reached by this processor with the Intro Kit was, here is the answer.
http://valid.x86-secret.com/show_oc?id=86036
Having reached up to 335 FSB, though in a totally unstable way.
|
