AMD AM3 Platform TestI got the AMD Phenom X3 720 BE from Massman to play with. Sadly this CPU is not one of our stronger candidates. We tried unlocking the 4th core on several motherboard’s (hence why I bricked the MSI GD70 by updating the bios to a beta version), but this CPU was very reluctant to do so. Secondly it wasn't one of the better overclockers either, topping at 3.4-3.6 depending on cooling used. Luckily for you, I got the second [M] motherboard, but had to promise not to do any flashing on it. The DFI DK 790FXB-M3G5 would serve as motherboard. More info on the guinea pigs here in Massman's review :
AM3 motherboard shootout: DFI DK 790FXB-M3H5 vs MSI 790FX-GD70
System ready for the tests with Push and pull fan installedAll hardware was installed in the same case as the Intel socket 775 setup, the Antec P160. Note that this case was premodded before to install a dual rad in the top. Since this case serves no longer as a watercooled equiped system the two 120mms fans provide extra flow of air in the case.
To clarify what we got for normal air ; 4 X 120mm case fans installed. This is 1 x front fan, sucking fresh air into the case, 1 x rear fan, blowing hot air out. And 2 x 120mm fans doing the same via the top of the case. All the 4 fans are 120mm 1000rpm case fans. When we install the Corsair H50 unit the rear fan is replaced by the 1600rpm Corsair fan, sucking in fresh air straight over the rad. The rest of the fans remain operative, only for the first test with the stock H50 test, we disabled both top fans. Then we are counting on the integrated fan on my OCZ power supply to act as exhaust. Second test was with the two fans enabled again.
Testing is done in the same manner as with the Intel CPU's. At first max CPU overclock is detected by exploring the limitations of the boxed cooler. Eg the I7 topped at 3.8Gz with the boxed cooler, even though the H50 could handle it at 4-4.2Ghz, 3.8 will be the maximum speed I'll test it on. This AMD CPU really had a hard time at 3.4Ghz with the boxed AMD cooler. 3400Mhz at 1.4Vcore it will be then, as booting into windows at 3.5 was barely doable, but temps were going nuts and stability was far fetched. 3.6 didn't even get me into windows, BSOD galore...
Before we move onto the results I still need to tell you that I couldn't mount the Thermalright eXtreme in the appropriate direction of airflow. I had only one choice and that was to mount it horizontally.
Thermalright TRUE installed, sadly only one way to mount the fan. For the ones that spotted it, the case backplates were all reinstalled before the testing began.I just had moved the videocard to the next PCI-E slot Secondly the DFI motherboard seemed to have issues with my GTX 285 when the card was in the first PCIe slot . Sometimes it would loose signal to the monitor and only a hard reset made me have an image again. Moving the video card down to the lower PCIe slot, solved that issue. Also this helped the TRUE's performance to be better as it created more space between the fan and the video card.
Performance results up next ->
I know the waterblock is copper , but the radiator is aluminium.