Original Link: https://www.anandtech.com/show/155
Epox MVP3C-M MVP3 Super7 Board
by Anand Lal Shimpi on April 28, 1998 2:19 PM EST- Posted in
- Motherboards
If you had asked the question, what motherboard would be the first to really work at a 100MHz bus speed back in December of last year most responses would have probably been centered around the motherboards that were unveiled at Comdex just a few months before. We expected the SiS 5591 chipset found on the MTech R581-A to be the first, however we were left with a disappointment. Elitegroup attempted to pick up where MTech left off, with their P5SD-B, unfortunately they could do no better than a reliable 90MHz setting.
One of the manufacturers that did the original VP3 chipset from VIA justice last year was Epox, their P55-VP3 was easily a knock-out upon its release. The Easy Setting Single Jumper setup made the absence of Jumperless Configuration Utility less of a pain. This time around, Epox is armed with a new chipset, the VIA MVP3, a new Jumper Setup, and even more importantly...a working 100MHz bus speed setting. In essence, Epox has the world's first example of a Working Super7 Motherboard on their hands.
Anand Tech Report Card Rating
95/A
CPU Interface | Socket-7 |
Chipset | VIA MVP3 |
L2 Cache | 512KB |
Form Factor | AT (w/ AT & ATX P/S Connectors) |
Bus Speeds | 60 / 66 / 75 / 83 / 100 MHz |
Clock Multipliers | 1.5x / 2.0x / 2.5x / 3.0x / 3.5x / 4.0x / 4.5x / 5.0x |
Voltages Supported | 2.1v / 2.2v / 2.8v / 2.9v / 3.2v |
Memory Slots | 3 168pin DIMM Slots (EDO/SDRAM) |
Expansion Slots | 1 AGP Slot 4 PCI Slots 3 ISA Slots (1 Shared / 1 Full Length) |
BIOS | AWARD PnP BIOS |
Taking a look at the layout of the MVP3C-M board, Epox did
a tremendous job with the design of this AT form factor product. While managing to
keep the size of the board relatively small, Epox managed to fit 4 PCI slots, 3 ISA, and 1
AGP slot on the board in addition to the 3 DIMM slots for memory expansion. Two of
the PCI slots will accept full length cards due to the positioning of the CPU Socket, a
socket which accepts and supports all current Socket-7 processors, including the upcoming
AMD K6-3D. From the looks of the chipset on the MVP3C-M, it would appear to the untrained eye that Epox's newest motherboard was nothing more than an excellent VP3 board, however upon closer inspection the VT82C598AT markings on the CPU to PCI bridge chip itself make it clear that this board means business, as it is a true Super7 motherboard based on the new VIA MVP3 chipset. |
|
The 512KB of 5ns L2 cache on-board is provided through the use of one chip complementing the 8ns Tag RAM. Due to the size of the L2 Cache in combination with that of the Tag RAM, the cacheable memory area of the MVP3C-M is 128MB, twice that of Intel's TX chipset. Upon boot-up the standard Award BIOS found on the board makes this clear by listing the cacheable DRAM size after displaying your system configuration information. Accented by the fairly thorough Epox manual, the MVP3C-M manages to make your initial introduction to the world of Epox products a pleasant one at the least, courtesy of their traditional "user friendly" features. At the top of this list is the newly revised Easy Setting Dual Jumper (ESDJ) CPU Setup which easily replaces the need for a Jumperless CPU Setup. Using a single jumper, especially the attractively colored green jumpers with elongated handles Epox provides for those "hard to reach areas," you can select the clock multiplier from a set ranging from 2.0x - 5.0x. Moving up the CPU Configuration Jumper Block on the board you can use another jumper to select the bus speed, ranging from 60 to a fully operational 100MHz setting.
|
One of the most highly anticipated features of the MVP3 chipset is the ability to run your SDRAM at a Bus Clock of 66MHz pseudo-synchronously while your CPU/L2 cache makes use of the 100MHz bus speed. This opens up the Super7 market to the low cost arena as well, allowing upgrade addicts to re-use their older SDRAM that may not be PC100 Compliant in newer motherboards while not limiting themselves to < 100MHz bus speeds. This setting is featured on the MVP3C-M, unfortunately due to the nature of the revision of the board tested by AnandTech (Revision 0.2) both the SDRAM and AGP clocks could not operate at 66MHz, rendering the AGP bus virtually inoperable at the 100MHz setting and rendering the system paralyzed if you attempt to make use of the pseudo-synchronous SDRAM clock which is derived from the AGP Clock. Epox has informed AnandTech that they have corrected this problem, so expect to see more from them in the coming weeks.
The quality of the MVP3C-M couldn't be better for a board of this caliber, not only does Epox package the board with the latest AGP Gart and VIA Bus Mastering Drivers but they also paid special attention to the engineering qualities of their newest pride and joy. Their attention to detail is represented by the rock-solid stability, especially when overclocked, and not to mention when using the previously unstable 100MHz bus speed. During the extensive testing that was performed with the MVP3C-M, the test system never once exhibited any sort of erratic behavior while operating under even the most extreme conditions, where most motherboards would crack under pressure, the MVP3C-M continued to function flawlessly Naturally, the performance of this motherboard using the 100MHz Front Side Bus Frequency is unparalleled by anything out today, simply using the 100MHz bus (which worked fine on both Intel and AMD processors; the Cyrix 6x86MX tested was a MX-200+ which is notorious for not being an overclocker's chip, and therefore didn't make it to the lowest setting of 100 x 2.0) increases overall system performance by a factor of 8 - 13% in Business Applications and Games. But before you get too excited, there are a few things that need to be fixed with the MVP3C-M before it makes its official release...let's take a look at those before making the final decision on this board. |
Don't worry about the AT form factor this board is available in, within a couple weeks time AnandTech will be receiving the ATX version of this board with a full Megabyte of Level 2 cache, that will double the maximum cacheable memory limit of the board to 256MB.
The only real problem encountered with the MVP3C-M was the fact that the board would not divide the AGP Clock properly at 100MHz bus speeds, making the use of a PCI Video card almost required for stable operation. Luckily Epox has informed AnandTech that they are working on this and it will be fixed by the time the board is officially released. Also, the lack of "in-between" voltage settings separating the 2.2v from the 2.8v Vcore settings on the board make it difficult to break the 333MHz mark with the K6-300 (3.5 x 100 isn't stable enough, most likely due to the voltage). Other than that, Epox has a definite hit on their hands.
IRQ Usage
Allows user to individually set IRQs for each Legacy ISA card
Epox MVP3C-M Chipset Features Setup |
||||
Item | Recommended Settings | |||
FPM | EDO/SDRAM 60/66MHz Bus | EDO/SDRAM 75/83/100MHz Bus | Safe | |
Bank 0/1 DRAM Timing: | 60 ns | Turbo | Turbo | Normal |
Bank 2/3 DRAM Timing: | 60 ns | Turbo | Turbo | Normal |
Bank 4/5 DRAM Timing: | 60 ns | Turbo | Turbo | Normal |
SDRAM Cycle Length: | 3 | 2 | 2 | 3 |
SDRAM Bank Interleave: | Disabled | 4 Bank | 4 Bank | Disabled |
DRAM Read Pipeline: | Enabled | Enabled | Enabled | Disabled |
Sustained 3T Write: | Enabled | Enabled | Enabled | Disabled |
Cache Rd+CPU Wt Pipeline: | Enabled | Enabled | Enabled | Enabled |
Read Around write: | Enabled | Enabled | Enabled | Enabled |
Cache Timing: | Fast | Fast | Fast | Fast |
Video BIOS Cacheable: | Enabled | Enabled | Enabled | Disabled |
System BIOS Cacheable: | Enabled | Enabled | Enabled | Disabled |
Memory Hole At 15Mb Addr.: | Disabled | Disabled | Disabled | Disabled |
Aperture Size: | 64M | 64M | 64M | 8M |
AGP Transfer Mode: | 2x | 2x | 2x | 1x |
Cyrix M2 ADS# Delay: | Enabled | Enabled | Enabled | Disabled |
Recommended SDRAM: Corsair PC100 SDRAM
SDRAM Tested: 1 x 64MB Corsair PC100 SDRAM
Manufacturer: Corsair Microsystems
Purchase Web-Site: http://www.nf-ny.com
The Test
In recent times, choosing a motherboard cannot be completely determined by a Winstone score. Now, many boards come within one Winstone point of each other and therefore the need to benchmark boards against each other falls. Therefore you shouldn't base your decision entirely on the benchmarks you see here, but also on the technical features and advantages of this particular board, seeing as that will probably make the greatest difference in your overall experience.
How I Tested
Each benchmark was run a minimum of 2 times and a maximum of 5 times, if the motherboard failed to complete a single test within the 5 allocated test runs the OS/Software was re-installed on a freshly formatted Hard Drive and the BIOS settings were adjusted to prevent the test from failing again. All such encounters were noted at the exact time of their occurrence.
Business Winstone 98 was run at each individually tested clock speed, if reliable scores were achieved with the first two test runs of the suite an average of the two was taken and recorded as the final score at that clock speed. If the test system displayed erratic behavior while the tests were running or the results were incredibly low/high the tests were re-run up to 5 times and an average of all the test runs was taken and recorded at the final score at that clock speed
After each motherboard was tested a complete format of the test hard drive was initiated and the OS/benchmarking software was re-installed afterwards a defragment was initiated using Windows 95's Disk Defragmentation Utility
Tests using AGP Video cards were only run under Winstone 97, as the AGP Millennium II does not affect the Business Winstone 98 score when compared to the PCI Millennium II used in the tests.
No foreign drivers were present in the test system other than those required for the system to function to the best of its ability
All foreign installation files were moved to a separate partition during the test as to prevent them from effecting the test results
All tests were conducted at 800 x 600 x 256 colors
Test Configuration |
|
Processor(s): | AMD K6/233 ANR AMD K6/266 AFR AMD K6/300 AFR Intel Pentium MMX 233 Cyrix 6x86MX-PR2/200 |
RAM: | 1 - 64MB Corsair PC100 SDRAM DIMM |
Hard Drive(s): | Western Digital Caviar AC21600H |
Video Card(s): | Matrox Millennium II (4MB WRAM - AGP) |
Bus Master Drivers: | VIA 2.14 |
Video Drivers: | MGA Millennium 4.07.00.700 VIA Gart VxD |
Operation System(s): | Windows 95 Service Release 2.1 |
Ziff Davis Winstone - Windows 95 Performance |
|||
Winstone 98 | Winstone 97 | ||
Business | Business | High End | |
AMD K6/200 - 100MHz x 2.0 | --- | 57.4 | --- |
AMD K6/291 - 83MHz x 3.5 | --- | 62.6 | --- |
AMD K6/300 - 66MHz x 4.5 | 19.4 | 60.2 | --- |
AMD K6/300 - 100MHz x 3.0 | 21.3 | 66.2 | 31.0 |
AMD K6/333 - 83MHz x 4.0 | 21.1 | 64.8 | 30.6 |
Intel Pentium MMX 200 - 100MHz x 2.0 | --- | 54.5 | --- |
Intel Pentium MMX 250 - 100MHz x 2.5 | --- | 58.5 | --- |
Once Epox perfects this Superpower of the Mainboard World it will be an excellent complement to the AMD K6 or even an older Pentium MMX, in any case, Epox will have another job well done and another recommendation from AnandTech under their belts by the time this baby ships. Let's see how quickly it takes the competition to respond...Shuttle for one, has a pretty hot product on their hands, the HOT-591P (MVP3 AT) that might boast equal if not better performance than the Epox board, unfortunately the layout has yet to be improved. Once again, we are left to play the waiting game.