Thermal Conditions during Work
Now let’s pass over to some practical tests of our Shuttle SN85G4 system. As I have already mentioned the cooling system is one of the most important elements of any SFF PC, and its efficiency is the key to the product’s success in the market. All components used in Shuttle SN85G4 dissipate quite a bit of heat: the CPU dissipates a lot, no doubt, besides, you will hardly use a RADEON 9000 in a system like that. Therefore, the cooling system will have to work really hard. Luckily the NVIDIA nForce3 150 chipset designed with 0.15micron technology dissipates very little heat.
We will test the efficiency of our cooling system in the following way. At first we will definitely assemble our barebone: install AMD Athlon 64 3200+ (working at the actual frequency of 2GHz, as you remember), RADEON 9700 Pro graphics card, two Corsair XMS3200C2 memory modules and Seagate Barracuda ATA IV HDD. Then we will install Windows XP Professional SP1 and Motherboard Monitor utility, which will help us track the CPU and system temperature and fan rotation speed throughout the test. After that we will do the actual testing.
For our test we will take 3DMark2001 SE and a distributed computing project client called Find-a-Drug, which we will run in the background mode and which will load the CPU up to 100%, without interrupting the 3DMark test. 3DMark will be set to perform 5 tests in a row, which should warm up the system quite nicely. As soon as the tests are over we will check how hot the CPU and the system became, and how high was the maximum fan rotation speed (we have already mentioned that it varies depending on the temperature). For a more illustrative comparison I am going to provide the same measurements for the system in idle mode. So let’s have a look:
Max. CPU temperature
Max. System temperature
Max. fan rotation speed
As you can clearly see, the cooling system is going just great: the fan hasn’t reached its maximum rotation speed a single time during the entire test session (it should switch to the maximum rpm when the temperature exceeds 60oC). It means that you are very unlikely to have the fan rotating at its maximum speed during the regular system operation, which also means that the noise level will be really low (this evaluation as “low” is definitely my subjective opinion). Of course, the above described results do not tend to be considered a universal truth valid for all cases, because the environmental temperature may as well affect the results of this experiment. However, overall this is a very efficient and well-done cooling system.