AOpen AK77-8X Max 18th January 2003
This is the second of the AOpen boards we received just before Christmas and is based on the Via Apollo KT400 chipset. Like the AX4PE Max it comes in a stylish black PCB color and all the extra connectors required including no less than three back panels which are often left out by other manufacturers to reduce cost. The drawback of this cost cutting is that it becomes almost impossible to obtain the back panels on their own. For example, Abit don't usually provide an extra back panel to use all the available USB ports and these are not stocked by retailers so have to be obtained directly from Abit which is lengthy and works out more expensive. It's nice to see everything included in the box with the AK77-8X Max.
The Serial ATA connectors are worth accumulating as these drives become more popular (actual availability is required first though) over the course of this year.
The board itself is feature packed. Here's a rundown of the manufacturer's specifications:
More information on any of these features can be obtained by clicking on the above hyperlinks. We will be covering Silent BIOS and Silent Tek in a separate article as part of our series on quieter PCs. The manual was thick and comprehensive although the black and white photos of a motherboard with a black PCB design made it quite difficult to locate components. It would have been better to use block diagrams to show jumper positions etc. without cluttering the diagram. The large full color fold-out poster was quite adequate for anyone to install the board without referring to the manual and was a useful inclusion.
Doctor Voice is a great benefit when diagnosing POST errors (as are the diagnostic indicator lights on the board itself) and gives audio error messages rather than forcing the user to interpret a sequence of long and short beeps which has been the norm for many years. In beeper mode the quality is very poor and sounds like a railway station PA system on a bad day. When the jumper is set to speaker mode (and connected to external speakers) a clear voice states the error in a strong American accented voice. We tested several different error types and the voice was correct in each case.
The AOpen jukebox allows the user to enter a special BIOS mode where the PC just acts as a CD/MP3 player without booting up. In conjunction with the Silent BIOS feature it is possible to actually turn off the CPU and case fans (since the CPU is not actually doing much) so there is no background noise to detract from the music. It certainly is an interesting feature for those that don't have a separate CD player.
The DieHard BIOS is very useful to restore the main BIOS if it gets corrupted (through over-clocking etc) and changing a jumper will restore the BIOS to a working state.
One quirk we did notice was the requirement to set FSB speed by moving jumpers on the board. Although the BIOS allows 1MHz increment changes these are up and down from the default FSB speed set by the jumpers. Speeds supported are 100, 133, 166 and even 200MHz should AMD ever release an Athlon based chip that has a 200MHz FSB.
The settings are quite flexible and if you turn off unnecessary things (like disabling the S-ATA controller since there wont be any drives for a while yet) the boot time into XP is lightning fast.
Test Set Up
To test this board we used a Pentium4 2.8 GHz CPU with 2 sticks of Crucial PC2700 memory and a Creative Ti4600 graphics card. We also used TwinMos PC3200 memory but quickly switched permanently to PC2700 for the reasons given below. As with the AX4PE Max we were unable to use CAS2 with any memory we tried let alone the setting for CAS level 1.5 which the BIOS provides so all tests were run at CAS2.5 and incidentally we could find no difference between loading the default and performance BIOS presets.
Boot time was very good once XP was installed and we could use the Vivid BIOS feature to put our own logo up during the POST phase. One up and running we ran the usual SiSoft and Mad Onion benchmarks along with DIvx tests. The latter may sound like a CPU test but is actually more a function of bandwidth and latency efficiency. Here's what we found.
We ran CPU Arithmetic, CPU Multimedia, Cache Memory and Memory Bandwidth benchmarks.
As expected no problems with making full use of the CPU.
Again within expected parameters.
A good result.
Here we see the shortcoming of the KT400 chipset against the NForce2 chipset. The dual channel controller on the NForce2 makes more efficient use of the memory than does the KT400. Using DDR400 gives results as shown which are good but still synthetic. Why didn't we use the DDR400 in all of our tests? The answer is below.
3DMark and PCMark
We have the PC2700 memory results on the left and the PC3200 memory results on the right.
As we feared the results of using the faster memory give us less performance. This is one of the annoying things we have come to expect from VIA where they release a chipset with inadequate testing and rely on members of the public to find all the bugs so they can correct them with a future chipset revision. A prime example was the KT266 and the revised KT266A. The problem is down to using memory speeds that are not synchronous with the CPU FSB speeds. It would not be a problem if every 333MHz FSB Athlon could over-clock to 400MHz FSB but this is extremely rare. The rest of our tests are conducted only with PC2700 memory.
The results are good for a KT400 board but not as good as an NForce2 board.
The performance of the board with gaming benchmarks needs to be determined so we ran UT2003 which will have to do until Doom3 becomes available.
For some reason we have what appears to be some kind of CPU limitation at the lower resolutions although the results are good at the higher ones due to optimized AGP bus settings.
We see the same puzzling results echoed here.
For consistency we will use Jet Li's The One as our test matter. It is not interlaced and contains a mixture of action types and is not too long. There will be three tests all using Divx 5.02. Audio will be encoded separately. I will try and keep my commentary to a minimum as all configuration information is shown in the images below.
Firstly we will start with the industry standard Xmpeg with its settings at those recommended by Intel and as used by other sites such as Toms Hardware. Here are the CODEC settings:
and here are the Xmpeg settings:
and here are the results
The performance is quite poor compared to the NForce2 board and this test really does stress FSB limitations.
We ripped our source material to hard disk and created a DVD2AVI project file using forced film (it was 99% film). Loading this into Gordian Knot we first saved an .avs file with no changes at all (720x480) and no filters of any sort. This was loaded into VirtualDub with the following CODEC parameters :
After encoding we got these results:
Again, slower than the NForce2 board.
This is all good stuff but how about a real-world test? To simulate a realistic test we added a neutral bicubic resize filter in the .avs file and used the following CODEC parameters (including two popular Pro settings) which are designed to total 700MB (when the audio is muxed in):
Which resulted in the following.
This test is more CPU intensive but we are still lagging behind the NForce2 board.
What about audio? We took the AC3 track from the above sample material and used HeadAC3he to convert it into Vorbis format so our final muxed file could have Ogg containment. There isn't space here to go into the advantages of Ogg Vorbis over MP3 and AVI so let's just say that Vorbis sounds about the same as MP3 for half the file size or twice as good for the same file size (that is subjective though).
These are the settings we used:
Since it is more meaningful to show throughput than time taken (which depends on the length of the source) we display the results thus:
A very good performance here from the AK&&-8X Max. There are no bandwidth limits stressed by this test and it is just down to optimized use of the CPU and we have a 20% boost over the NForce2.
What can we say? Despite the good audio encoding performance the overall impression of the AK77-8X Max is that VIA have missed the boat in the competitive stakes. This is no fault of AOpen and the board is up to their usual high standards and packed with all the extra features we have come to expect from their Max series. There really is no compelling reason to buy a KT400 chipset these days unless you are short of cash or only have one DIMM (therefore being unable to take advantage of Dual Channel DDR). If that is the case you can pick up an AK77-333 for 2/3 the price of this board and much cheaper than an NForce2 board. If you're looking for a top of the line Athlon board then we can't quite recommend the AK77-8X Max as an all-round performer.
We would like to thank AOpen Europe for the review sample motherboard.
All trademarks are the property of their respective owners.