Aopen AX64 Pro Motherboard Evaluation

Pages: 1 2 3 4

Test Results

Visual Inspection

Probably the first thing most people will notice is that the board has only 4 PCI slots, which seems to be a big issue for many DIY’ers. Most would happily give up the AMR slot that is also present for an extra PCI slot. There is also, of course, an AGP slot. Up to 1.5GB of SDRAM (or VCSDRAM) can be installed into the 3 DIMM slots, however, due to the lack of a memory buffer, chip count becomes an issue. AOpen recommends that modules should not have more than 16 chips (which is generally good advice in any case). The now-standard headers for Wake-on-LAN, Wake-on-Modem and IrDA cables are also present, as well as two front USB headers. The floppy and IDE connectors are located in front of the DIMM slots, putting them right next to the drive bays, and the ATX power plug is along the edge of the board between the DIMM and CPU slots.

The standard PC99 compliant ATX connectors are present (parallel, serial, USB,PS/2 mouse and keyboard), as well as a game port and RCA connectors for audio. The latter are due to the AC97 audio capability provided by the on-board CODEC chip, however this can be disabled via jumper (JP12).

As with all Aopen boards, the BIOS setup includes a ‘Turbo Defaults’ setting for optimum performance, and has multi-language support (German, Chinese, Japanese and English). The Host (CPU) clock speed is user selectable, with 66.8, 75, 83.3, 100, 105, 110, 112, 115, 120, 124, 133, 140 and 150MHz options. CPU ratios from 1.5x to 8.0x are also available for those with non-locked processors.

There is one particular aspect of the clock settings that can be somewhat confusing for the uninitiated, which is the variation in PCI, AGP and DRAM speeds. This is of particular interest to overclockers, since it can cause problems with devices that cannot tolerate certain clock frequencies.

The PCI clock frequency has been called ‘asynchronous’ on some review sites, however this is incorrect. An asynchronous PCI clock would always run at 33MHz, regardless of the Host clock frequency. In reality, the PCI clock is derived by dividing the FSB speed by a ‘PCI clock ratio’ of 2, 3 or 4.

The range of FSB speeds that are available in the BIOS is determined by the setting of JP23 and JP29. The ‘default’ jumper setting will cause the BIOS to detect the default FSB speed for the processor installed. FSB speeds between 66MHz and 83MHz use a PCI clock ratio of 2, those between 100MHz and 124MHz use a ratio of 3 and speeds from 133MHz use a ratio of 4. This means that a FSB speed of 83.3MHz would result in a PCI clock speed of 41.66MHz, while 150MHz Host clock would result in a 37.5MHz PCI clock, etc. Pages 15 and 16 of the manual provide the details of the various jumper settings and resulting speeds.


Verification – As part of the standard methodology, I first tested the board with QuickTech Pro 2000 and RAM Stress Test diagnostics, from Ultra-X. QuickTech Pro is a system level diagnostic and burn-in tool which allows us to verify that the board has no obvious defects that would prevent a reliable evaluation (see our review of this software for more information). RAM Stress Test is a very intensive memory diagnostic that allows us to verify that any stability issues are not due to data integrity problems because of defective memory or bus timings that are too fast (refer to our review for details of this product). With all three DIMM slots populated, we ran RST through 10 loops without encountering any errors. QuickTech Pro was then run in burn-in mode for 8 hours, again with no errors reported.

The next set of tests involved our standard circuit-level diagnostics using the PHD Plus and PHD PCI diagnostic cards, also from Ultra-X. These diagnostic cards measure the signals for all circuits (IRQ and DMA controllers, PCI and memory busses, system timer, etc.) against reference timings, and if there is even a small deviation, a ‘failure’ is reported. These tests were run 100 times in succession with each card to make sure there were no intermittent errors.

Generally, AOpen motherboards will not have any errors at all with these tests, however this board did. DMA Page Register tests failed about half of the time, while the Interrupt (channels 0-7) and DMA transfer tests failed consistently. These tests are specific to the ISA bus, and I did not test any ISA cards in this board. The results of this circuit level test would indicate that caution should be exercised if you intend to use an ISA card in the motherboard. Though the results don’t necessarily mean your ISA card will not work, you might try to get your vendor to verify that your card will work without any compatibility issues, if possible. Another possible source of information might be the AOpen newsgroup, which we have provided a link to under the ‘Resources and Test Notes’ section.

Stability and Reliability – I first installed Windows 98 (second edition), and ran the Passmark Burn-in test. This is a multi-threaded burnin program that puts a heavy load on various aspects of the system simultaneously, including the processor, graphics and I/O subsystems. The Burn-in test was run for over 24 hours without any errors reported. I also installed and ran Business Winstone 99 and Content Creation 2000 tests in Demo mode for another 8 hours each. The test settings were set to stop on any error, and to reboot after every run. In speaking with ZDBOp, this is the only reliable way to use demo mode as a ‘stress test’ because is eliminates the problems caused by memory leaks in the applications used by the benchmark. There were zero errors reported during this 24-hour test period.

I next installed Windows 2000, and ran the Passmark burn-in test for 72 hours. Winstone 99 (Business and High-End) and Content Creation 2000 were run in demo mode for 20 iterations without rebooting (about 4 hours each) without any errors being detected.

The AX64 Pro appears to be as stable as any motherboard recently tested, however it is important to note that these tests were not run using an AGP 4x video card, which currently seems to be where most compatibility issues are being reported. For business use, using a 2x AGP video card, I can give this motherboard a very high recommendation for stability. With features such as the Die-Hard BIOS, batteryless operation and resettable keyboard fuse, this is certainly a motherboard to consider for those desiring reliability as well.

Pages: « Prev   1 2 3 4   Next »

Be the first to discuss this article!