Asrock X48TurboTwins-WiFi

Asrock manages to scare Josh Collins, but for all the wrong reasons.

Asrock is well known for supplying the market with cheap, cost efficient budget-grade motherboards.

It’s been responsible for Frankensteining some of the most polar opposite technologies onto the same board, too. In the past the company has produced motherboards with both AGP and PCI-Express graphics slots and even a board with two different CPU sockets, by supply an add-in card for those who really desperately crave options.

With this history, it is no surprise to see DDR2 and DDR3 on the same board. Then again, a number of manufacturers have been providing this to the market, such as the big players GIGABYTE, ASUS and MSI.

Asrock is now, more than ever, competing with some of the largest manufacturers in the world. Surprising to some and unsurprising to others, the company has decided to do battle by making a shift towards a focus on overclocking.

Those keeping their overclocking ear to the ground would have heard of Biostar’s recent rebirth as the producer of an overclocking monster, the TPower I45, which appears to rather effortlessly pass 600MHz FSB and even 700MHz FSB.

While the Biostar board does have the advantage of the new P45 chipset, what the situation as a whole proves is the ability for a small player in the manufacturing business to provide a rather stunning board. Unfortunately for the peeps at Asrock, its recent efforts have left much to be desired.

To put a board through a gruelling test scheme, it first needs the OS, drivers and applications installed. To do this, we load the optimised defaults within the BIOS to assure the utmost stability. Sadly, the crew of BIOS engineers at Asrock seem to have had their silly caps on when they set some of these optimised values.

The optimised defaults had settings such as the LAN, 1394a FireWire and HD audio all disabled. Furthermore the primary GPU is set as PCI rather than the far more common PCI-Express and, perhaps the most frustrating of all, the USB controller is disabled; this effectively includes all USB legacy support. As a result, even though you’ve booted into the BIOS and set the optical drive as primary boot device, if you’re using a USB keyboard you won’t be able to hit a key to start the installation of the OS! We hope the BIOS engineers slap themselves if/when they read this.

After finally getting the OS installed, we set out to optimise the platform wherever possible. Unfortunately, when operating with a 333MHz FSB, the maximum memory frequency possible through a divider is DDR3-1333. Furthermore, this divider doesn’t even allow the platform to boot into the OS when voltage options are set to ‘auto’. This meant for the first bout of testing we were restricted to a lowly DDR3-1066 – we may as well have been using DDR2 memory!

This could have been rectified with some voltage tweaking but to say this motherboard left a big ‘WTF!?’ stamped on our forehead is an understatement when it comes to the motherboard voltages.

Anyone used to using any motherboard of at least some decent quality, and with a clue about system setup, would want to know exactly what values the system voltages are at. Regrettably, the BIOS engineers at Asrock believe values of ‘auto’, ‘low’, ‘middle’, ‘high’ and ‘highest’ are appropriate for settings such as NB vcore, SB vcore and VTT. This in itself was the single reason why we would not overclock this motherboard – who knows how many volts these settings put through to your precious memory and CPU!

Further BIOS level discrepancies include the GTLREF being decided by a decimal value multiplied by the VTT. This in itself is fine but as already mentioned, we have no way in knowing the VTT voltage.

This also leads to another issue; we can’t even check NB vcore, SB vcore, VTT or vDIMM because the BIOS only allows voltage monitoring for CPU vcore, 12v rail, 5v rail and 3.3v rail – whether these are accurately calibrated is also questionable.

Pair this BIOS-level misery with a horrendous motherboard layout – just check the spacing between the PCI-Express slots or the positioning of two of the SATA ports and the ATX 24-pin connection if you don’t agree – and you’ll understand why we advise you to avoid this motherboard.

We’re yet to get a street price for this Asrock board in the Australian market but considering there are solidly featured, well laid out options at a number of price points from the likes of ASUS and GIGABYTE that also have a better engineered BIOS, the only redeeming feature for the Asrock is the fact it uses solid state capacitors. But then, so do the ASUS and GIGABYTE options.