Upon first glance, the FA-11 appears to be a no nonsense Socket 370 design. The board comes with a standard complement of 1 AGP, 5 PCI and 1 ISA slots (the ISA slot is a mfg. option). No AMR slots are present and the optional embedded PCI audio support doesn’t appear to be stuffed on the US release version. The features that are present represent the bare essentials and that isn’t a bad thing at all. Most users that integrate DIY systems won’t use an AMR slot anyway, and typically prefer the option of choosing their own brand of audio card. Currently the market has seen a rise in motherboard component pricing, so keeping the components down to the essentials makes a lot of sense. The typical cost for this board, from the Pricewatch search engine, is around $100.
The FA-11 comes equipped with three DIMM slots that can handle up to 1.5GB of either SDRAM or Virtual Channel Memory (VCM) SDRAM. The Apollo Pro chipset allows the FSB (Front Side Bus) and DRAM clock to run asynchronous, which means that at a FSB speed of 133MHz, it’s possible to run the memory at HCLK – 33MHz (i.e., 100MHz). The memory handling difference between the FA-11 and the BX based FB-11 is ironic in that even though the FB-11 has an extra DIMM slot for a total of 4, it can only house a total of 1GB. This stems from the fact that the Apollo 133A chipset can handle 512MB DIMM modules and BX designs are limited to 256MB modules.
It was nice to see an FIC design that took up a reasonably small footprint in comparison to the extremely large AZ11 and SD11 board designs. This board should fit into most cases large or small without much issue. The placement of the various board connectors seemed very reasonably to me, since the IDE and floppy drive interfaces are strategically located in the upper right portion of the board. There a few dip switches for setting up bus speeds and multipliers that might be slightly harder to access once all the cables are plugged and the board is installed, but this is a very minor observance.
In this day and age of most board settings available in BIOS settings, the FA-11 is a throw back design with a 6-position rocker dip. With this 6-position switch it’s possible to set CPU multiplier ratio, clear CMOS and clear password. Right next to the rocker switch is a three-position jumper block to set CPU type (Intel or Cyrix) and front side bus speeds. In the opposite corner of the board are two more jumpers for enabling over voltage settings (called JOCK) and Keyboard Wakeup. Fortunately FIC has provided very legible silkscreen legends for the various switch settings so that you don’t have to dig out the manual every time you make a change.
To verify that the board functioned adequately under stress, I first loaded up the FA-11 with my PIII 550Mhz with the board set to 5.5 x 100MHz FSB. Loading up Windows98SE and I began with the BCM memory stress test while having two of the three DIMM slots loaded with 64MB of PC133 modules (for 128MB total). The BCM memory test consists of a number of complex test patterns written and read back from the memory. Testing was completed overnight with over 200 iterations of the complete memory test with no failures or problems were found.
I increased the FSB setting to 133MHz and then began running my 550MHz PIII at 733MHz. Using the utility program DIMMTEST, I began looking into the FA-11’s ability to run the memory asynchronously. One of the touted features of this board is the ability to run the memory modules at speeds different from the Host clock. In the BIOS section, under the heading of DRAM clock, there are three choices: AUTO, Host CLK and HCLK – 33M. Unfortunately no matter what setting I used, the DIMMCLK program reported 100MHz. I’m assuming that a BIOS revision may be in order to correct this shortcoming. I will provide an update if and when FIC finds a fix for this problem.
Both the Ziff-Davis Winstone 99 and Content Creation 2000 were loaded and each was run in the demo mode 10 times consecutively with a reboot in between each run. Scores showed that this board is certainly a good implementation of the VIA Apollo 133A chipset and seemed to stand up to some of the better boards I’ve tested. At 733MHz (5.5x133MHz), the scores from both Winstone 99 and CC2000 turned out to be practically the same as the system running the PIII 850MHz (8.5x100MHz FSB). This just goes to show you what having a faster FSB speed can achieve compared to slight increases in CPU speeds.
The Intel Celeron II 566MHz was installed next without a hitch. The BIOS displayed it correctly at a Celeron II 566MHz and it booted up fine. The numbers of the benchmarks were certainly lower than the 733 or 850MHz PIIIs, but things were stable and didn’t hiccup at all. The 66MHz FSB of the Celeron II really does slow things down. The PIII 850MHz processor was briefly used to verify proper operation through some CC2000 and WS99 testing and it worked flawlessly.
Passmark’s Burn In Test was used over about a week’s time while I allowed the board to run days on end out on my test bench. The Burn In Test is a multi-threaded stress test that’s capable of testing your CPU, Hard drives, CD ROMs, Sound cards, 2D graphics, 3D graphics, RAM, Network connections and Printers. Using Burn In Test, I put the FA-11 through a long-term stability test using the 733MHz Pentium III (5.5x133MHz FSB). Most of my testing was done with the ambient temperature frequently over 100 degrees F and the board functioned without a problem. Using the board’s standard settings the board exhibited exceptional stability.
Be the first to discuss this article!