EPoX MVP3F-A Review
To say this board is unusual is a bit of an understatement. Because it is unusual I am not going to do the typical review of it but rather point you to the Epox 58MVP3C-M for its performance marks because in reality they are virtually the same.
The MVP3F-A is a Micro-ATX board of 243x220mm (about an inch shorter that the MVP3C-M AT board, with an incorporated 512k L2 pipelined burst cache. What makes it so unusual is that it has a ridiculously few expansion slots and onboard Yamaha YMF715B sound. The board has the requisite AGP slot and VIA's MVP3 chipset allowing it to run the front side bus at the 100MHz speed required to make the board an official Super7 mainboard but is limited to 2 PCI slots and 2 ISA slots and to make matters worse 1 of each is a shared slot, effectively making available only 3 expansion slots. I am still kind of baffled as to this mainboard's place in what is primarily an upgrade market. I suppose that Epox thought that the use of AGP and onboard sound would reduce the need for expansion but I feel as though they went a bit too far. If however you have a niece or nephew that hasn't a clue about what's happening inside a computer and has no desire to learn, you could incorporate the board into a desktop or mini-tower case and produce a low cost gaming and study machine that would make a nifty Xmas gift.
The board sent to me by Epox for review has a lot of the features that make their mainboards a pleasure to work with. The ESDJ (Easy Setting Dual Jumpers) allow you to set the CPU frequency (this board lacks the 95MHz CPU frequency - Default setting for the AMD K6-2 333MHz) and clock rate by simply adjusting a pair of clearly marked jumpers while another 10 pin block easily lets you choose the proper voltage setting for all of the current socket 7 processors. You are severely limited in this area to core voltages of 2.1v, 2.2v, 2.8v, 2.9v, and 3.2v which tends to limit this board's ability to overclock somewhat. With the use of the MVP3 chipset the memory can be adjusted to run at the CPU bus clock or fixed at 66MHz by the implementation of VCS (Virtual Clock Synchronization) technology using a delay lock loop to enable synchronous or pseudo-synchronous operation of the CPU, AGP DRAM and PCI buses. This makes it possible to run the board with SDRAM DIMMs that do not meet the PC100 standard while running the CPU frequency at 100MHz. The board makes provision for 2 DIMMs but no SIMM slots are incorporated, so you will not be able to use your older Fast-Page or EDO SIMMs as you can on many of the current Super7 mainboards available. The board supports up to 256MB of system memory but with the L2 cache size at 512k only 128MB of that memory is cacheable and exceeding this limit could slow down performance.
The product is packaged along with an excellent easy to follow manual, a pair of diskettes containing the latest AGP and Ultra-DMA Bus Master drivers as well as the driver CD for the onboard Yamaha sound. Setting up was a breeze. Although an admittedly small mainboard all of the connectors and jumpers have been efficiently placed to allow easy access and pulling all of the components I used off of the Soyo SY-5EH I am testing concurrently I was a little concerned that the proprietary ETEQ version of the VIA chipset might have caused some device and driver conflict but was happily suprised that it booted right up replacing only the PCI Host Bridge in the process. Since I was using the ATI Expert@Work AGP card I went ahead and installed the included AGP driver just to be on the safe side and Windows 98 called for its installation CD to gather the necessary drivers for the onboard Yamaha sound. These I promptly replaced with the drivers on the included audio drivers CD. In all of the testing I did only one glitch occurred. When switching processors from the Cyrix MII 300 to the AMD K6-2 300 something strange occurred. A first run-through of Winstone 98 produced unusually low scores and after a little detective work I found that even though the BIOS had correctly identified the processor's frequency it was only running at 150MHz. Rechecking all of the jumpers and rebooting I used WinTune 98's CPU test to see if any changes had occurred but once again the processor was only running at 150MHz. I tried a number of different settings and each time the processor only ran at half the speed reported in the BIOS startup screen. The problem mysteriously disappeared when I changed processors to an Intel P55C 200MMX and I thought I was using a defective processor but, tossing it into another machine brought up the expected 300MHz CPU frequency and when I put it back into the MVP3F-a it ran just fine. I have no explanation for this occurrence but thought it significant enough to mention.