Testing

Once those small problems –created mostly by our anxiety - were overcome, we began the testing. Like we always do, the comparison shown in the charts will be made against a cooler stock. Although this may look absolutely excessive -and as a matter of fact it is-, the cooler stock is the only cooler every user has with no exception, and to which anyone could easily run this comparison test. If you want to see the results of other cooling solutions, just check our other reviews. ;)    

Like in previous reviews, the tests were run with 32M Super PI while the evolution of the temperature was controlled during the entire process, before the beginning of the test until it stabilized to the temperature mark prior to the test. None of the habitual tasks, such as the e-mail client, anti-virus, etc., were interrupted to make the CPU work as hard as possible.  

For the first round of tests, a Panaflo fan TCO3BZ-1BX, FBA12G12W model (120mm, 12V, 0.45A) was used due to its very low noise level.


To get a first approach to the behavior of the Ultra-120, we ran a test with the regular configuration of the CPU when using its cooler stock at stock frequency and barely 1.2 V. The result was too obvious for this solution: 2º C difference between IDLE and FULL LOAD, which is more than you can expect from any cooler, but let's be realistic, no one would consider such a heat sink just to keep the CPU at regular conditions.  So we tested the CPU again, this time under the same operating conditions we always test coolers, 300 X 9, 1.66 V, which is excessive, but as we wanted to raise deliberately the CPU’s temperature, we thought it was perfect.

It is worth mentioning that –as always- all the tests are run at 24º C room temperature.

Hmm, things started to get interesting!!! When the CPU was subjected to very high voltage and significant operating frequency we got as a result, an outstanding performance: just 14º C difference between the room temperature and the CPU at FULL LOAD (maximum temperature reached by the CPU: 38ºC), which was close to the temperature obtained with the cooler stock but in IDLE mode. It is not necessary to mention what happens to the CPU with the cooler stock at FULL LOAD, where the temperatures go beyond the acceptable values for a continuous operation.

It could be said already that we are looking at a top cooler device, and not only are we referring to its size, but also to its efficacy that puts it above most of air coolers, as well as several liquid cooling systems.

As we enjoy doing this so much, like most of our readers, we decided to take the tests a bit further and experiment with different configurations and fans.

To begin with the new set of tests, the silent Panaflo was replaced with an “unbearable" Delta TFB1212GHE  12V, 2.94A; 220CFM and a noise level that practically makes you wear ear plugs if wanted to spend some time near one.   

Next, we ran the tests under the same conditions. After the three usual runs to avoid possible errors –three for each new configuration-, we checked that the temperature had dropped slightly -less than 1º C- when compared with the previous results (this was far from the 2 or 3º C we speculated before running the tests). 

At first we were surprised, but as we moved forward with the tests and the reasons for the obtained results were getting clearer, it was hard not to become a true fan of the Ultra-120.

It is also important to mention that from the moment the CPU went from FULL LOAD to IDLE until the temperature returned to the original value, the elapsed time was less than 5 seconds. This shows clearly the incredible capacity of the Ultra-120 to extract the heat from the CPU.

We continued with three more “configurations". For them we used two equal fans of medium quality, but with a lower noise level than the Panaflo. One thing that called our attention about the Ultra-120 is that it didn’t include two pairs of fan wire clips, as we thought that adding a second fan would improve its performance. It didn't take long for us to realize how wrong we were.

The configurations were the following:

1^st configuration: one fan only
2^nd configuration: two fans, one on each side (a fan wire clip was improvised).
3^rd configuration: same as the previous one but with an “enclosure” attached in order to cover the outer parts of the fines and therefore, prevent the air from escaping, and complete the entire trajectory from one end to the other.

Just as with the results obtained with the Delta, all three results were placed in a narrow band of less than 2º C and with a maximum temperature that was of 39º C, barely higher than that obtained with the Panaflo. This confirmed what we had been suspecting since the previous test: our interpretation is that due to the efficient design of the fines, it is not necessary for the Ultra-120 to have a high capacity fan to carry out the task. Consequently, this could justify the reason for not having included another fan wire clip for a second fan, as what is beyond a top quality and low noise fan only brings problems when assembling it.

We believe it's better to add that second fan to help the case's overall ventilation than wasting it in a place where it would not do much to drop the CPU's temperature.        

However, it is possible that in configurations in which a greater generation of heat is produced, the assistance provided by a higher air flow achieves better results, but in this particular case that was analyzed, the performance of the fines was enough to keep the heat pipes’ cold end cooled. 

It is highly advisable for each user to find the appropriate configuration and fan for his system. Any speed regulator would help you find this.