Quote Originally Posted by Kosior View Post
The IFX is alot worse than TRUE.... The only difference is - if I'm correct - lack of that small backside cooler

so... where could be the problem ?:>

Also - what about mounting 3 fans on IFX ? would it give an enormous difference or not?
I've been comparing the IFX with and without the backside HS to the TRUE ever since the IFX was released. I've tested open bench and in two different cases. Many cases are too cramped to fit the backside HS. I compared them in a full sized Armor that is equipped with the 25cm side door fan and also in a Lian Li PC60B PlusII.

The PC60B is not the best breathing case out there by any means. It does have a 12cm front inlet fan and a 12cm inlet in the side door but the exhausts are just 8cm fans on the back and out the blowhole. With the case fans in the stock configuration, the TRUE and IFX had almost identical performance, and different fan configurations on the IFX had little effect. The Zalman 9700 was at most 1C warmer than the TRUE and IFX in the PC60B with the fans set up stock and often the temps were identical. That made me think that the case airflow was masking the performance differences between the HSFs. So I took the front fan and flipped it around to an exhaust, used a Scythe Kama Bay to add another 12cm front exhaust and put in a higher flow side door intake fan. That dropped temps for all HSFs by 3 to 5C under load. With the IFX set up with fans in pull/pull orientation, it outperformed the TRUE by 1-2C and the 9700 by 2-3C.

The above led me to pull the guts of my image server that lived in the Armor. My Armor has two Icages and therefore two 12cm front intake fans. It also has a 12cm rear exhaust, a 9cm rear exhast and a 9cm blowhole fan in addition to the 25cm side door fan. After checking all of the fan cfm specs, I flipped the two front fans around so they would be exhausts, mirroring what I'd done on the PC60B. It worked great. The mobo, GPU, NB and CPU temps all went down by ~5C below the lowest PC60B temps. The gap between the TRUE and the IFX widened but the 9700 got closer to the TRUE. Without the rear IFX HS, the IFX was 1 to 3C cooler than the TRUE and the 9700. With the backside cooler on, the gap was 2 to 3C. Again, I ran the IFX in pull/pull and I think that works well with a side door fan that pumps air into the gap between the two TS towers.

So I yanked the rig back out of the Armor and compared the three HSFs on open bench. If I just used the fans that were on the HSFs, the temps were equal to or 1C higher than those in the Armor with the same ambients. That was surprising but maybe it means that the Armor exhaust fans and the big door fan really add to the HS airflow picture. So I put a four high flow 12cm fans around the CPU HSFs and they brought the open bench temps down to where they beat the Armor results. Still though, the IFX had an ~2C advantage over the TRUE and 9700.

My first comparisons were done on a DFI 3200CFX with a S939 Opteron 185 running at 3GHz. But most were on an Asus Blitz Formula with a Q6700 at 3.4GHz with no water on the NB Fusion block.