More testing :

I tried some gaming on the IGP, which is based on ATI's 4290 chip. ASrock recommends a max resolution of 1024 x 768. I can heartily recommend that as anything above becomes very choppy. Of course kid’s games like Freddy Fish and co will run effortlessly, but shooters, sports games etc will require a separate GPU to warrant fluid game play. IGP's have come a long way and are improving daily. We wonder if more sideport memory ( eg 256-512Mb ) wouldn't make it the ultra killer IGP mobo.

Here's eg Far Cry 2 at 1024 x 768 Resolution with Low and Medium detail settings :



Even at medium detail FarCry2 got very laggy, dipping below 24Fps is never a good thing. The low detail setting remained pretty smooth during the benchmark tests. These tests were conducted at the stock GPU clocks with UMA set at 512Mb.

Once you start to push the IGP, for which there are separate tweaking settings like GPU and Sideport RAM speeds and voltages. The sideport RAM speed got critical over 1333Mhz. Setting 1400Mhz resulted in graphical corruption. Here's a screen from the HWbot April IGP competition. CPU on LN2, IGP on air at 1.55GPu voltage. 1050Mhz is where it all stopped on air, more Mhz and volts lead to lockups.



Putting LN2 on the IGP resulted in the HwBot winning 36386 score, setting new standards for IGPs. We will devote a complete article to what we have done and tried to get there....

UCC testing went flawless with the 555 BE cpu, By enabling the UCC in the bios we got directly 4 cores in the windows. Without any tweaking these were rocks table during a few hours prime testing.



In the above picture you can also spot ASrock's own OC tool. Very handy when trying to boost performance or monitor voltages from within Windows. It's not as elaborate as AMD's Overdrive, but it has got the necessary tools for a quick and easy setup.



USB3.0 was tested by using the Sunbeamtech Airbox 3.0 external HD bay. We used HDtune to measure performance and also copied several gigabytes from the external USB3.0 HD to another. With transfer rates averaging at 60Mb/s. It was almost twice as fast as when using the same drive hooked up to the USB2.0 ports (30.1Mb/s)



Thuban testing however didn't go as smooth as other tests. Maybe the latest Beta bios I got ( L1.51) isn't the correct one (yet) for the new AMD X6 CPU. The X6 was directly recognised, but OC'ing remained problematic. Turning off the Turbo function helped improving stability. Yet Vcore was jumping all over the place. Setting 1.4 in the bios resulted in 1.43 in Windows and over 1.47 when stressing the CPU. To avoid overvolting my sample I stayed around 1.35 in the bios. Probably a few corrections will enhance compatibility with the new Thuban series CPU's.