Case Design and Picking a CPU Cooler

[Computer-Ed]

When I first started asking questions about this I thought the answers would be easy to come by and already done, I was surprised. My question was would a case that uses a CPU vent be better off getting only a standard top down cooler instead of the vertical cooler? In contrast would the vertical cooler be better in a directed front to back only air flow system.

I was really surprised as I interviewed CPU cooler companies for our radio show that NO ONE had a real answer for this. I had planned to test this with a Thermaltake V1 and my Big Typhoon VX but circumstances got in the way, to the rescue came Zalman with a 9700. Since the 9700 and the Big Typhoon tend to be in the same league for cooling I figure these will work for the comparison.

Now this is NOT a head to head comparison of coolers, this is a testing of style in case design. As such I am not going to spend this article sitting specific temperatures and getting lost in that debate, I am instead looking at temperature changes as we alter the parameters and see what has the most effect in case and heatsink design.

For the cases I am using the Antec Nine Hundred and Antec TX640B. To determine the raw effect of the vent I have turned off the fan in the back of the case and will use a PSU fan for the warm air exhaust. Remember we are not judging actually performance but the effect of the vent design on the two cooler designs.

For testing the CPU is an E6600 that is running in an overclocked state at 3200. The thermal compound is Arctic Silver Ceramique so I can do this without long setup times. The testing is being done at 100% load using ITAT and using the highest core temp after 30 minutes of run.


Test Series 1:
Vent Baseline: Vent is open and back fan is running on the TX640B case. This test was to provide the TX640B baseline numbers. In the base testing the VX was able to keep the CPU cooler by 3C.

Vent Only: In this test the back fan was turned off and the testing rerun. I am here looking for how much effect the back fan had on each CPU cooler and to see if the vent to down design of the VX could better hold temperature control. Both coolers saw a noticeable jump in CPU temperature but the VX jump only 7c compared to the 9700 jump of 12c.

Vent Closed: Next I closed off the CPU vent and turned back on the back fan. This would determine if the CPU vent was making any real difference in the base numbers for the 9700. The VX jumped 3c over the baseline while the 9700 only 1C.

Conclusions: In the case of a vented CPU design the top down does hold an advantage over a similarly powered vertical design. The vent allowed for a nice pull of cooler outside air even if the exhaust air flow was limited. However if the vent is shut off the pure exhaust system seems to favor the vertical design which is blowing its air straight at the exhaust, thus increasing its overall air flow.
Test Series 2:
Baseline: This series uses the Antec Nine Hundred case to see how much a strong front to back flow can affect the temperature. The systems are setup using the stock Nine Hundred configuration. The base line results saw a reversal of the previous baseline with the 9700 winning by 2C.

Just exhaust: I wanted this test of work only with the exhaust system and so I blocked off the side opening on the Nine Hundred case and turned off the front fans. The VX jumped by 7C and the 9700 by 3C.

Intake on low: next we nudge up the intake a bit to see what this does to both. The VX changed by being 5C above baseline the 9700 dropped to 1C above baseline.
Intake to medium: again we nudged the intake up and again the temps changed with the VX going to 3C above baseline but the 9700 held at baseline.

Intake on High: With the fans on high and the vent still closed the VX was still 1C above its baseline while the 9700 was still in the baseline range.

Now this test is far from scientific and I would love to see others try it to compare the results. With this information in hand however it would appear hat common sense is right and the case design has a fair impact on which cooler design to pick.

I think the design is impacted by the effect it can have on the air movement. By this I mean a top down design is meant to pull cool air and blow it down through the fins. If the case has a thermal vent the air is drawn directly from the outside and thus is cooler and more effective.
The vertical design however throws air back at the rear of the case pulling warmer air from inside the case. However if you have a powerful exhaust fan pulling air out of the case right behind you in effect turbo the air flow. Hence the reason the Zalman's smaller fan can move as much if not more air than the VX in a proper design. (Could not test this but the numbers seem to indicate it.)

What does all this mean? If you are building a system with the primary air flow is forward to back AND it has no CPU vent then the stand up cooler with the fan forcing air toward an exhaust is the best solution. If you have a CPU vent then the top down design can really shine and should be seriously considered.

This testing is not conclusive by any means but I think the base numbers are interesting.