Page 3 of 3 [ Jump to Page: 1 2 3 ]
Warning: Always look at the date when you read a hardware article. Some of the content in this article is most likely out of date, as it was written on December 12, 2011. Check out our more recent articles.
We've spent a lot of time fine-tuning the airflow in the Antec P183 V3, so in order to give the P280 its due diligence we will be spending a fair amount of time trying out different fan configurations to find out which work the best. More fans do not necessarily mean better cooling, and since we are always very concerned about the noise levels of our systems we want to have just the right number of fans installed in just the right locations.
We could just show the end results of our testing, but in the spirit of transparency (and since we see no reason to not publish information we spent the time and effort to gather) we will show all of our temperature results. If you do not want to see our various fan configuration testing, feel free to skip ahead to the next section where we will compare the cooling of this chassis to the Antec P183 V3.
To test the cooling performance of this chassis, we will be using the following hardware:
To fully load the system, we will run a combination of Prime95 and Furmark until the temperatures stabilized (roughly 10 minutes). This will cause higher temperatures than any normal user is likely to see, but should give us a clear worse-case scenario for cooling. Temperature readings were taken from the hottest core of the CPU (Core 0 on our test CPU), the video card(s), the ICH (southbridge) and the IOH (northbridge). The margin of error for our temperature results is 1°C.
Keep in mind that this was a learning process, so as we discovered what fan configurations worked and which didn't, we adjusted the locations of the fans for future testing accordingly. For that reason, our first configuration (the Gelid Tranquillo Rev2 with a single GPU) has more fan configurations than our later configurations.
Gelid Tranquillo Rev2 - Single GPU
Right away, our first test configuration with the Gelid Tranquillo Rev2 and a single GPU gives us some interesting results. The default fan locations (2 top and one rear) gave us the worse load CPU temperatures by 2°C, but was among the best for our motherboard and GPU temperatures. Adding a single fan to the middle of the chassis improved further on the motherboard and GPU temperatures while also bring the CPU load temperature in line with the other fan setups. Overall, the default fan locations should work just fine for most system configurations, but an additional fan located in the middle mount is recommended for especially hot configurations.
Another interesting point is that adding a fan to the mounts in the very front of the chassis helped our CPU temperatures a bit, but overall made our motherboard and GPU temperatures worse. It's odd to think of a chassis doing better without a front intake fan, but this chassis has been optimized by Antec to operate under negative pressure (more exhaust fans than intake) which these results fully support.
Gelid Tranquillo Rev2 - SLI GPU
With the same CPU cooler and two video cards in SLI, we see an expected overall rise in temperatures. Video cards and CPU's have different thermal thresholds and while we still are not at all close to the manufacture’s thermal limits, we are beginning to approach our comfortable limit of 80°C on the CPU and 90°C on the video cards. You will notice that we did not do any testing without two fans in the top of the chassis. This is due to the fact that our best results from the previous test all had both top fan locations populated.
Once again, we feel that having the two top mounts, the rear mount and one extra fan in the middle mount is the best fan configuration. Interestingly, having two fans in the middle fan mounts, or having one in the front and one in the middle, results in overall higher load temperatures than just a single fan in the middle. The one degree drop in GPU temperature (which technically is within our margin of error) by having the second fan in the middle is frankly not worth the rise in CPU and motherboard temperatures.
Coolit Eco II - Single GPU
With the Coolit Eco II cooler, the airflow in the chassis is much different so we were not sure what to expect. We expected the lower CPU temperatures, but we were surprised to see a drop in video card temperatures. Another thing that we did not expect was how consistent the temperatures are across the different fan configurations. All of the load results were within 2°C of each other, which makes deciding which configuration is the best a bit difficult. The two top configurations are the green (2 top, 1 mid, 1 front) and the red (2 top, 1 mid). The differences between the two are within our margin of error, so we are going to call the red configuration (2 top, 1 mid) the winner since less fans means less noise.
Coolit Eco II - SLI GPU
With two video cards in SLI, we see a bit more variance in temperatures, but only on the ICH. All of the other temperature differences are within our margin of error. Since we have a second video card to worry about, we tried a bit different of a configuration this time with one and two fans in the middle fan mount. Both of those configurations were our best performers, but once again the difference between them was within our margin of error. Just like the last test, we give the red (2 top, 1 mid) the winning bid.
Now that we know that the ideal cooling setup for the P280 is to have the two top fans installed, a rear fan (if available) and a single fan installed in the middle of the chassis, we can take a closer look at how the cooling in this chassis compares to the Antec P183 V3. We will be using the exact same hardware, simply transplanted into the Antec P183 V3.
For cooling in the P183 V3, we will be using the same Antec TwoCool fans from the P280. We want to have the same amount of potential airflow, and using the Antec TriCool fans that come with the P183 V3 would give it an unfair advantage in terms of cooling. We will be installing the fans into the locations that we know from experience will give us the best cooling. For the Gelid Tranquillo Rev2, this will mean having fans in both of the front intakes as well as the rear exhaust. For the Coolit Eco II, we will keep both of the front intake fans but move the rear fan to the top mount (still as an exhaust).
We would normally never run this hot of video cards in SLI in the Antec P183 V3 without adding a custom side fan, but since the Antec P280 does not come with a side fan mount we will not be adding a side fan to the Antec P183 V3. Because of this, we expect high temperatures on the video cards when in SLI but we want to see how these two chassis fare against each other without any case modifications. To help give you an idea of how substantial the difference a side panel fan makes, we recommend reading our Side Panel Fans: Are They Worth It? article.
In this round of testing, we will also be taking thermal images in addition to our sensor readings to help us identify any hotspots that do not show up on our temperature logs.
For our first configuration (Gelid Tranquillo Rev2 with a single GTX 580) we see quite an improvement in temperatures with the P280. Surprisingly, the CPU temperatures are identical but the motherboard (ICH and IOH) and GPU temperatures are much better in the P280 with about a 5°C reduction in temperature. The thermal images show almost the exact same data, although it is interesting to see that the bottom half of the motherboard is hotter in the P183. This is due to the fact that the P280 has a fan just to the right of the GPU, while the closest fan in the P183 V3 is all the way at the front of the chassis.
Adding a second video card gives us some results we were not expecting. This time, only the IOH and ICH sensors show substantial difference in cooling between the two chassis. While the GPU temperature readings do not show much difference, the thermal image shows us that the video cards in the P280 are in fact running much cooler. The video cards in the P183 V3 are much hotter across the entire GPU even if the temperature of the core itself is roughly the same as in the P280. This goes to show that while thermal sensors are great most of the time, they do not always show the full picture.
Moving on to the Coolit Eco II CPU cooler, the motherboard is overall hotter than with the Tranquillo Rev2, but this is to be expected since the CPU fan is no longer in the center of the chassis. We again see the same ~5°C reduction in IOH, ICH, and GPU temperatures that we saw in our Gelid Tranquillo Rev2 with a single GPU testing. The CPU temperatures are identical, but since the CPU cooler is now always getting fresh air from the rear of the chassis, we do not expect any appreciable changes in CPU temperature. Overall, these results are essentially the same as our first Gelid Tranquillo Rev2 configuration.
In our final configuration, we again see the same relative results that we saw in the Gelid Tranquillo Rev2 with the video cards in SLI. The motherboard and GPU run much cooler in the P280, while the CPU temperatures are identical. The hotter area around the CPU and between the video cards in the P183 V3 is also worth pointing out.
Overall, we are impressed with the cooling of the P280. Especially considering that Antec puts an emphasis on quiet operation in the Performance line of chassis, the P280 cools better than we expected. Our thermal testing also shows that most system configurations will be just fine with the stock Antec TwoCool fans in the default locations. Users who need more cooling can easily do so however by adding one or two fans to the interior fan mounts.
Cooling is not the only selling point of this chassis however; quiet operation is also a major component. For that reason we also took acoustic measurements of each configuration in both the Antec P183 V3 and the Antec P280 in addition to our thermal readings.
Our acoustic readings were taken 0.6m (~24 inches) from the front-top of the chassis which is the ISO 7779 computer noise standard's defined "Seated User Position". This puts the microphone at about the same position as the end user if the chassis is on the ground next to a desk. Our ambient noise level is 27.5 dBa and since we do not have a perfectly isolated environment, our margin of error is a fairly large .5 dBa.
Starting with the Gelid Tranquillo Rev2 CPU cooler, the first thing to notice is how much quieter the P183 V3 is at idle. 2-3 dBa may not seem like much, but it is definitely noticeable in person. Under full load, we get a mixed set of results. With a single GPU, the P183 V3 is still quieter, but only by .6 dBa. With two GPU's in SLI, the P280 is technically quieter, although the results are close enough to fall well within our margin of error.
With the Coolit Eco II CPU cooler, we again see much better noise levels at idle with the Antec P183 V3. Interestingly, our load results are switched from the previous configuration. With a single GPU, the P183 V3 and P280 are close enough together to be within our margin of error. Adding the second GPU in SLI is what allows the P183 V3 to show how much the additional acoustic dampening available in that chassis helps with noise levels.
So what conclusions can we make from this data? First, an inherently quiet configuration (Core i5 CPU, passive GPU, SSD or Western Digital Green hard drive) or a more powerful configuration (like our test configuration) at idle will be quieter in the P183 V3 than in the P280. A louder configuration like our test configuration at load or something approaching a server-class configuration (multiple CPU's, additional cooling needed for chipset) will be just as quiet in the P280 as in the P183 V3.
Our cooling performance testing in the previous section thankfully goes hand-in-hand with our noise results. Inherently quieter configurations generally require less cooling, so while the Antec P183 V3 does have slightly worse cooling than the P280, it is still more than adequate for those configurations. More powerful configurations on the other hand can really take advantage of the extra cooling in the P280 while still maintaining roughly the same noise levels as the P183 V3.
The Antec P280 is a great chassis for a wide range of system configurations but really shines in keeping high-end configurations running both cool and quiet. Lower power configurations may run quieter in other chassis such as the Antec P183 V3, but since Antec is touting this chassis as a high-performance option this is not much of a surprise. The fans that come with the P280 are great, although the amount of airflow is low enough that we would recommend replacing them with Antec TriCool or similarly rated fans on configurations involving SLI.
We would really like to offer this chassis in our product line as an option for high-end configurations but there are just enough issues to keep us from doing so. Our biggest issue is with the internal fan mounts which our testing has shown to be critical for keeping high-end configurations running cool. Unfortunately, they are designed in a way that would not fare well when the system is shipped across the country.
Our second issue is in regards to the front LED/switch cables. They just barely reached on our test motherboard which means that on many motherboards the cables would have to be strung across the motherboard just to get to the headers. In such a well-designed chassis from such a reputable company, this is a major oversight.
Until these issues are resolved we unfortunately are not able to add this chassis to our product line. We are working with Antec to resolve these major problems and hope to be able to announce an update soon.