Introduction

Our upcoming series of µATX articles has traveled a long road (Ed: that's an understatement!). When we first envisioned a long-overdue look at the µATX form factor motherboards, we thought it would be your typical motherboard roundup with maybe a twist or two tossed in to keep it interesting. One thing led to another and before you knew it, our minds started to run rampant with additional items that we felt were important for the article. This led to scope creep and those of us who manage projects - or who have been unlucky enough to be on a project that has featuritis - know what happens next.

That's right, we over-emphasized the new article features to the detriment of our primary focus, providing a motherboard roundup that featured the often ignored but market leading µATX form factor. What started out with adding a couple of features such as IGP video quality comparisons and midrange CPU performance turned into a maze of thoughts and ideas that led us to believe it would be quite easy to add additional tests without affecting the overall schedule too much. We were wrong, but we hope that our future motherboard articles will be better for it.

How did we get stuck in the quagmire of µATX hell? It began with innocent thoughts of adding budget to midrange CPU coverage, low to midrange graphics comparisons against the IGP solutions, High Definition playback comparisons utilizing not one but each competing standard, Windows XP versus Vista versus Linux, onboard audio versus add-in cards, and even tests of input devices and external storage items. It ended with our project scope changing from being motherboard specific to platform encompassing.

We started down that path but despite periodic excitement, at times we also ended up with a dreaded case of paralysis by analysis syndrome. Don't get us wrong: we do not regret the effort that has been expended on this roundup; however, we sincerely regret the time it has taken to complete it and we apologize to those of you who have been waiting months for this information. It turns out that we ignored one of our favorite quotes from C. Gordon Bell, "The cheapest, fastest, and most reliable components are those that aren't there." That is one of the many factors that caused us problems, as it became quite obvious during testing that getting all of this equipment to work together and then benchmarking as planned was not exactly going to be a walk in the park.

We have been constantly waiting on that one BIOS or driver to fix a malady of problems that we've discovered along the way. The manufacturers would ask - sometimes plead - for us to retest or wait as "that problem is being solved and a fix should be available immediately". Immediately it turns means days and weeks, not hours. We also received several product revisions during the course of testing that required us to throw out the old results and start again. In the end, we hope our efforts paid off and at least we have the knowledge that every supplier has had ample opportunity to fix any ills with their product.

Our experiences with a wide variety of components will be discussed extensively in a series of articles to be published over the coming month. However, at the end of the day, the star of this show is still the motherboard. If the CPU is the brain of a computer and the video card is its eyes, then the motherboard is the central nervous system. It truly is the central focal point of the system and having one that works correctly makes it really easy to put a system together.

As such, we are changing our testing emphasis from being primarily performance based to a combination of performance, features, stability, support, and those intangibles that we experience during testing that might set one board apart from another. While performance is important, does a few tenths of second or an additional two frames per second in a benchmark really mean that much when you cannot get a USB port working due to a crappy BIOS release or your system does not properly recover from S3 sleep state when you are set to record the last episode of the Sopranos? We thought as much also, so we are changing our vantage point on motherboard testing.

While we are performance enthusiasts at heart, the fastest board available is not worth much if the included features do not work as advertised or the board constantly crashes when trying to use an application. Our testing emphasis, especially between boards based on the same chipset, will be focused on stability and compatibility with a wide range of peripherals in both stock and overclocked conditions. Speaking of features, we will place a renewed emphasis on networking, storage, memory, and audio performance. More importantly, we will provide additional analysis on overclocking, energy consumption, cooling capabilities, layout, and power management features where applicable.

We also want to take this opportunity to put the manufacturers on notice: we will not countenance delays, patches, and numerous updates again, particularly on products that are available in the retail market! If a lemon of a motherboard gets released to consumers and it needs more BIOS tuning or perhaps an entirely new revision, we are going to do our best to point this fact out to the readers. We understand that it can be difficult to get every single peripheral to work properly, especially with new devices coming out all the time, but when a motherboard fails to work properly with a large number of USB devices, memory types, GPUs, etc. that product shouldn't be on the market.

At the end of this journey we will provide three different platform recommendations based on the various components we have utilized in testing. Our platforms are designed around HTPC, Gaming, and Home/Office centric configurations with a heavy emphasis on the systems being quiet, reliable, and affordable. Okay, we blew the budget on the HTPC configuration but we will provide several alternatives to help control costs on that particular buildup. Let's find out what else is changing and exactly what will be included in our comprehensive review of the µATX motherboards and surrounding technologies.

Benchmarking, GPUs, and Displays
Comments Locked

42 Comments

View All Comments

  • Mazen - Monday, August 27, 2007 - link

    If memory serves me well, we were supposed to see something last week!
  • Mikus42 - Thursday, August 16, 2007 - link

    You have great timing! I am researching components for a Micro ATX box.

    I have an older Shuttle SFF I am going to replace. (It is topped out on CPU speed.)

    For me, noise is a very important factor. My shuttle is on my desktop and it can be a tad noisy.
  • jonp - Thursday, August 9, 2007 - link

    The SG03 case is $160 (Newegg) w/o a power supply....ouch.
    The SX6 w/600w case is $112 (eCost).

    I would like to nominate the Foxconn TLM776-CN300C-02 w/300w at $40 (Newegg) as a more typical mini-tower case option.
    2 5.25" external -- one more than SG03; same as SX6.
    2 3.5" external -- one more than both SG03 and SX6.
    4 3.5" internal -- two more than both SG03 and SX6.
    expansion slots, front ports same for all three.
    power & reset buttons (SX6 yes; not sure if SG03 has reset button).
    1 80mm exhaust fan (SG03: 120mm intake, no exhaust; SX6 80mm intake, 120mm exhaust)
    Steel (SG03 Al; SX6 steel(?))

    The TLM776 power supply is a Channel Well unit. If that doesn't suit then it can be replaced with say a Seasonic S12 II 330watt at $60 for $100 total.
  • RamIt - Wednesday, August 8, 2007 - link

    I am so dissapointed by the numerous delays with this article, i will just pass it up when it finally materalizes. Way too many broken promises Gary.
    Peace out.
  • SanLouBlues - Wednesday, August 8, 2007 - link

    Since you're dipping your toes in, I'm dying to see some reviews of the PCHDTV cards for linux that claim ATSC/NTSC and unencrypted QAM support.
  • carver5678 - Wednesday, August 8, 2007 - link

    here are some suggestions for the review:

    (1) measure northbridge temps. *lots* of people are having problems with hot NBs. for example, see customer reviews on newegg product pages.

    (2) evaluate the ability of the boards to support passive cooling. can the scythe ninja fit on the motherboard? does the board rely on airflow from a cpu cooler to cool its northbridge? can it fit a scythe ninja, thermaright hr-05 NB heatsink, and thermaright hr-03 VGA heatsink, all at the same time, or do they conflict with each other?
  • lopri - Wednesday, August 8, 2007 - link

    I am 200% for the AT's (newer) stance on motherboard evaluations. Contrary to common myth, enthusiast market is HUGE. In various forums I often hear arguments like "Vendors make most money from OEM contracts", "They don't care about enthusiasts. They make money off server market", etc. - this can't be wrong enough. As computers are becoming a commodity, the number of enthusiasts is getting bigger, and vendors are very well aware of this fact. These days we observe even the most conservative and OEM-oriented companies attempt to jump in this growing market and have a piece.

    And there come the inevitable side-effects: Rushed-out (i.e. bug-fest) products, irrelevant blings, non-functional features, unacceptable compatibility, and poor longevity, etc.

    The #1 components in PC, of the said side-effects, are by far motherboards. I have been really put off by today's motherboards and for me it totally ruined the image of "Intel Stability". In all honesty, if I were to be responsible for an on-going support of an Intel based system, there would be only select a few 975X boards to choose from – of course in order to avoid potential hassles . It has been that bad, IMHO, and the ways vendors handle their mishaps are infuriating sometimes. Anyone who had to deal with the so-called "tech support" of famed motherboard manufacturers would know. Even under the best case of scenario, users have to deal with 2~3 weeks of downtime for replacement. (maybe except EVGA) For me, it's a total nonsense.

    If you buy a TV, you most likely expect it to work as advertised. Same for a DVD player, refrigerator, or anything that you spend your hard-earned cash on. Not so for a motherboard. It's been such a mystery for me (who happen to be dealing with many professionals and their works/products) how these Taiwanese mobo manufacturers can get away with the poor quality of products they dump in the market, and their shady business practices. When I purchase a product, I should be able to expect a 100% defect-free product, instead of praying on luck. I also don't want to see a almost same product as I bought, being sold with a slightly different name and fixed functions, not too long after my purchase.

    To a certain extent, I blame online journalist for their negligence. I do understand many are working with the manufacturers directly to get things right for us and that alone could be a reason why they often are sympathetic with the manufacturers. However, it is important to note that ultimately, reviewers should be in the buyer's shoes, and articles from reputable reviewers can have a huge effect on buyers' decisions. No matter how much work a mobo manufacturer put in, and no matter how much they listened to the reviewers, if the final products that end-users get are not up to the standard (read: 100% free from , reviewers should clearly state the shortcomings of the products at hand.

    Above said, I felt so refreshed reading from the first page of this article and would like to commend AT's courage. It's not that AT has been partial or unfair until today - it's the opposite. AT's motherboard reviews are probably the only motherboard reviews (and TechReport's) I've trusted to date and often times I don't even bother to read any other site's reviews. (they are waste of my time) I couldn't be happier with every single sentence that Gary wrote in a clear English. It read almost like a 'motherboard review manifesto 2007' :D and I couldn't agree more with every single word he stated.
  • SunLord - Wednesday, August 8, 2007 - link

    I sure as hell hope you have no plans to recommend the Mx5000 desktop to anyone. It's an unstable pile of crap. The keyboard at random times goes nuts and keys "stick" as you type. They don't stick down but they just keep repeating in windows cause the keyboard has dropped it's signal it's a bluetooth/site point problem. Oh andf can't forget the battery life gotta remember to always turn it off or you'll be replacing batteries every other week. The mouse is nice though.

    I'm buying a G15 and a rf wireless mouse to replace my month old Mx5000 because it's not worth the hassle of having to pull out and replug in the bt adapter every time the keyboard takes a dump
  • strikeback03 - Wednesday, August 8, 2007 - link

    We have a MX5000 set at work and I have one at home. On the work set it has never dropped the connection, and we have replaced the batteries twice in 15 months (and it sees use for a couple hours a day). The only major problem is that the Logitech software can't handle switching between Linux and Windows, which is why on my home computer I didn't install the Logitech software. I use my home keyboard more, and still get 3 months or so out of decent rechargable batteries.
  • SunLord - Wednesday, August 8, 2007 - link

    I exchanged one set after a week and this one is still flacky i just moved to linux 3days ago and while bluetooth is useless it works perfectly. So I'd have to say it's all logitechs fault with there typical craptastic drivers

    On batteries no clue I burned through 8batteries in a month i bought the same day i got the keyboard at BB... Nice new duracells

Log in

Don't have an account? Sign up now