Final Words

When Athlon 64 was first introduced, the motherboards that we saw were generally loaded with every option that could be crammed on the board. With the Athlon 64 sporting the memory controller on the CPU, manufacturers tried very hard to make their product stand out with features. There were a few notable exceptions, like the FIC K8-800T value board, but most of the boards were top-end with prices to match.

A lot has changed with Socket 754 since September 23, 2003. New, lower priced Athlon 64 CPU's like the 3000+ and 2800+ have appeared. The introduction of the 3400+ has also pushed the price of the 3200+ down to a much more affordable level. With the new Socket 939 to be introduced in May, we expect Socket 754 to become even more of a value board. This will be even more the case when mainstream Socket 939 products are introduced. In the end, Socket 754 will likely displace Socket A and become the bang-for-the-buck boards in the Athlon 64 family.

Why all this positioning? Simply because the AOpen AK86-L is the perfect board for the new Socket 754 reality. It is currently selling for below $100 on-line, and that certainly qualifies the AK86-L as a good value. Value, however, is more than price, and here, the AOpen AK86-L delivers like no board that we have evaluated to this point. Enthusiasts love a well-priced motherboard that they can take to the moon in performance, and the AOpen certainly qualifies here. With the latest 1.06M BIOS, you get a full assortment of tweaking options - including lower multipliers. Memory timing options are outstanding, and unlike most K8T800 boards that we have tested, the AOpen AK86-L is just as happy as you could want at 2-2-2-6 timings at DDR400. This makes the AK86-L a superb choice as a motherboard for that screaming Athlon 64 system. It even includes Gigabit LAN and SATA RAID in a board selling for around $100. The AK86-L was also an outstanding overclocker in our tests - with a processor that has been nothing special in the overclocking department.

Does this mean that the AOpen AK86-L is the perfect Athlon 64 motherboard? Unfortunately, it isn't - at least not yet. The board is faster than average, and a very good performer as a stock motherboard. It also has every tweaking option that we could want, but the ranges on some of the available options are just too limited. Memory voltage only goes to 2.7V where 2.9V to 3.0V would be much better; CPU voltage is also very limited at the top end, going from an excellent low of 0.8V to just 1.55V. The AMD stated limit of 1.7V would be preferred at the top. Fortunately for potential buyers, most of these "wish list" items can be accomplished with a BIOS update - if AOpen chooses to do that. The PCI/AGP operates on ratios and is not really a lock as we see on Intel boards, and for those of you who have to have Firewire, it is not available on the AK86-L.

Motherboards are often a world of cookie-cutter sameness. The AOpen AK86-L stands out in that world as a great value and wonderful performer. It comes just at the time when everyone is looking for value in Socket 754 boards. There is no board out there for the price that delivers the combination of features, flexibility, and performance offered by the AOpen AK86-L. If you are looking for a Socket 754 motherboard, the AOpen AK86-L is highly recommended. With a few more BIOS revisions, it could easily become our Editor's Choice among Athlon 64 motherboards.

Workstation Performance
Comments Locked

22 Comments

View All Comments

  • Pumpkinierre - Monday, February 16, 2004 - link

    In your update you say the new BIOS 1.06q includes cpu multipliers. But the old BIOS had that. So does this mean a larger range ie 0.5x divisions? Also now that you've got the Geiger, could you check the PCI lock at several clockspeeds?

    This is important to me as my interest in purchasing an a64 is to run it at high clockspeed (with only a mild o'clock) as I believe this would substanstially improve gaming performance. Of course this is also dependent on fast low latency DDR 466-500 being available- still yet to occur!.
  • Boonesmi - Thursday, February 12, 2004 - link

    what about agp/pci lock? have you been able to test it yet??
  • Wesley Fink - Thursday, February 12, 2004 - link

    AOpen has just sent a revised BIOS 1.06Q, which adds vDimm adjustments from 2.5 to 3.0V in .05V increments. This is an outstanding upgrade from the 2.7v max vDimm in the BIOS we tested and makes the AK86-L one of the best 754 boards available for memory overclocking.

    A kind soul is hosting the 1.06q BIOS and a link has been added to the review.
  • JTDC - Tuesday, February 10, 2004 - link

    Does anyone know where the 1.06 BIOS revision can be found? I cannot see it on the AOPEN site. Any suggestions as to where I can look?
  • peonyu - Friday, February 6, 2004 - link

    I have a Aopen board [not this one though,damn] and the options on it are awesome, same with its stability and performance. The only prob though is the same that this board has - VERY low voltage settings. For reasons i dont understand Aopen is very stubbern with allowing higher voltages, it happens with every board they release. So until they get past that and release a update which increases voltage I dont think it would be worth picking one of these up if your into overclocking.
  • Shinei - Thursday, February 5, 2004 - link

    Does Athlon 64 rely on timings as much as the Athlon XP did? I didn't think it did, considering the bandwidth it can use eclipses even the Pentium 4's...
    As for the article, now I really wish I'd waited a few months to upgrade, instead of going straight for the GA-7N400P2/2800+ in December...
  • Pumpkinierre - Thursday, February 5, 2004 - link

    It would have been nice if you had included some benchmark results at the different multiplier o'clocks or even standardised to 2gig (ie 9x222,8x250) now that you have an A64 board that doesnt have memory, multiplier, HT or (from initial observations) AGP/PCI lock issues. Naturally memory latency settings would have to be kept the same. My pet theory with the A64 is that increasing the RAM speed and lowering memory latency (which appears difficult on many a64 mobo/memory combinations) would be as good as a standard o'clock which to date has'nt been the a64's strong point. This seems like the right board but we're still waiting for CAS2 DDR500.
  • Boonesmi - Thursday, February 5, 2004 - link

    dang i really like that board.

    when are you expecting the "PCI Geiger"??
    hehe i want an update :)
  • Wesley Fink - Thursday, February 5, 2004 - link

    #1 - Four "3-pin" fan headers is correct, and the typo is fixed.

    #2 - We will be reviewing a few of the memory brands you mention in a few weeks. A socket 754 board will be included in the tests.
  • bhtooefr - Thursday, February 5, 2004 - link

    Damn, I want one! Of course, it would be even better if it were Micro ATX... (I've already got a mATX case, and it's very small)

Log in

Don't have an account? Sign up now