First Impressions

abit has done a wonderful job in the design and inclusion of high quality components on the AB9 QuadGT when compared to the AB9 Pro. The layout, feature set, performance, and BIOS options obviously place this board at or near the top of the Intel P965 product offerings. We really like the board and were impressed with its overall performance. In fact, the board just seems to be light years ahead of the AB9 Pro, but an otherwise excellent design is ruined by an immature BIOS. The issues we discovered during testing make us wonder if the light is on at abit's quality assurance department. We have briefly described some of the problems we faced during testing but there are more.

In our USB testing we found if the BIOS was changed from USB Keyboard or Mouse support via OS to BIOS then Windows XP would refuse to load if it was installed or if we tried a new installation from the CD drive. Additional USB testing revealed that any USB drive attached that is not a bootable device will result in the same behavior with XP not loading.

During storage drive testing we noticed the board would not run our Optical drives in DMA mode and defaulted to PIO mode regardless of how the JMicron controller was setup when utilizing the latest firmware and driver updates on our test image. We had to complete a new XP SP2 install with the JMicron controller setup as a RAID device in the BIOS and load the driver set via the F6 command in order to get DMA support for the Optical drives.

We also noticed IDE support from the JMicron controller was not available if we were using a RAID setup on the Intel ICH8R controller. We have noticed this before on other motherboards utilizing the JMicron chipset. However, this issue has generally been solved so we expect abit to have a solution shortly. For now, we recommend purchasing a SATA DVD drive and turning off the JMicron controller altogether unless you need the e-SATA support. If you use a SATA optical drive on the Intel controller the board will not recognized the drive if you have RAID enabled so this a catch 22 situation.

The JMicron fiascoes have hit every board manufacturer at one time or another and if you want to blame somebody then start at the doorstep of Intel for shipping a chipset without native IDE support before the market was ready for it. Of course, the motherboard suppliers could have opted for the VIA IDE controller that Biostar uses without issue, but then e-SATA or additional SATA port availability would not be on the option list. Considering most people would rather have at least one solid IDE controller rather than a couple more SATA ports that they often won't use, Biostar's decision seems to make the most sense, even if it doesn't look as good on the feature table comparisons.

Our retail sample was also delivered with the incorrect I/O panel shield so it appears to us that somebody was sleeping or watching Oprah during the packaging operation. Besides the issues listed above and the ones mentioned during the article we have to say the worst offense committed by abit is the memory voltage setting issue with the µGuru utility. The inability of the current µGuru application to correctly read and display the memory voltages set in the BIOS could directly lead to product failure as in our example.

Normally, we test each option and software switch and then reboot to ensure the BIOS setting matches the application in use. Unfortunately, after using version three of this application for the past few weeks we trusted it would work correctly. Instead of checking each setting before starting our benchmarks we assumed our memory voltage settings were correct after increasing our voltage slightly during the overclocking stability tests. We will no longer assume that µGuru and the abit BIOS are working in tandem before we start our test sessions. Needless to say, frying memory that costs over twice as much as the motherboard being reviewed was not a pleasant experience.

Overall, our initial experiences with the AB9 QuadGT have led us into a love and hate relationship with the board. We love the layout, performance, stability, and overall quality of the motherboard. This motherboard has tremendous potential and is a board we want to own for a long time. We hate the obviously early release of the product and the issues encountered to date as they should not be occurring. Several of the issues discovered in our initial testing can be blamed on an immature BIOS release and should be solved quickly, but we have to wonder why these issues even exist. It seems as if part of abit did not learn from the AB9 Pro launch while others took those same issues to heart and solved them. This situation reminds us of an old saying, "Those who do not remember the past are condemned to repeat it." After spending a few days with this product we can attest this saying has never been truer.

Benchmark Setup and Performance
Comments Locked

41 Comments

View All Comments

  • Operandi - Wednesday, January 24, 2007 - link

    Excellent.
  • Talcite - Monday, January 22, 2007 - link

    Do you guys mean 'Final Impressions' on the last page of the article?
  • Gary Key - Monday, January 22, 2007 - link

    It was First Impressions, slip of the tongue until we get the next BIOS release. ;)
  • mechBgon - Monday, January 22, 2007 - link

    Hey Gary, I think you got a typo:
    quote:

    Something else worth nothing is the audio features.
    You meant "noting," not "nothing," right?
  • JarredWalton - Monday, January 22, 2007 - link

    That was my fault, and it's what happens when you edit at 3 AM (and use voice recognition). Anyway, it's fixed, thanks.
  • sprockkets - Monday, January 22, 2007 - link

    Intel has such a good processor, then messes up on the IDE. I would gladly make a computer around such a chip, but the nForce 6150 and AMD/Ati chipset is much more compelling, and since my customer needs IDE anyhow, AMD wins.
  • MadAd - Tuesday, January 23, 2007 - link

    Its taken all this time to get rid of the parallel port yet the much more used IDE (which should have the same kind of legacy lifespan) is out the door in a flash.

    If I had a tin foil hat id be shouting conspiracy theory and wondering which one of the storage majors paid Intel off to pull a stunt like that.
  • Stele - Monday, January 22, 2007 - link

    quote:

    ...Japanese sourced solid state capacitors ...


    Oh no, that misnomer strikes again... capacitors which use solid aluminium electrolyte are properly called "solid aluminium electrolyte" or at least "solid aluminium"... there are no "solid state" capacitors since, in electronic parlance "solid state" is traditionally used to describe circuits that do not use vacuum tubes (electron flows in solid materials - the semiconductors within ICs - rather than vacuum/empty spaces such as in tubes).

    It's a minor point on the face of it, but Anandtech is well-respected throughout the IT community; hence a mistake or misconception could potentially reach and misinform a great deal of readers (and even writers) out there. :)
  • JarredWalton - Monday, January 22, 2007 - link

    Changed - not all of us are electronics experts. Solid state or solid electrolyte - is there a difference to the laymen of the world? ;)
  • Stele - Monday, January 22, 2007 - link

    quote:

    ...not all of us are electronics experts...

    Oh no, no, didn't mean it in any negative way; I understand that, so was just contributing what little I can to help keep Anandtech at the top :P

    You're probably quite right that the layman may not know the difference, and indeed they likely wouldn't care either way anyway. The ones we're worried about are those that are just entering the technical arena and/or are in their formative, learning stages... wouldn't want them to start on the wrong footing and spread the misstep when they in turn teach others ;)

Log in

Don't have an account? Sign up now