Biostar MB-8433UUD

Biostar had a US office in El Monte, CA but it is apparently no longer open. Biostar distributes in the USA through Asolid Computer Supply, their subsidiary in Fremont, CA. Biostar has WWW pages at biostar-usa.com in the USA, bioteq.com in Europe, www.biostar.net and www.biostar.com.tw. Cyrix lists Angelina Shiang (510)226-6678 as a Biostar contact.

There is at least one biotechnology company called BioStar in Boulder, CO, and Saskatchewan. www.biostar.com is the BioStar biotech company in Colorado, not the computer hardware manufacturer. Please don't bother them with motherboard questions!

• Biostar has a BBS at (818)443-4222 complete with BIOS updates for all of their motherboards. There's no index, though, so you just have to guess... You can see a capture of their file directory here. This lists a BIOS version of UUD960326i for the MB-8433UUD.

• Apart from this site and my Jumper Page, some information on Biostar's boards can be found at their WWW site at biostar-usa.com in the USA, bioteq.com in Europe, www.biostar.net (English) and www.biostar.com.tw (Chinese).
• Quantex Computers has a page of support information as well. They offer updated BIOS files for this motherboard on their web site. Try these at your own risk!
• Award, the maker of the motherboard's BIOS, says Unicore Software in North Andover, MA, provides updated BIOS for the Biostar boards. Check out Unicore's web site for more info.
• I do not work for Biostar and will not provide support information for this board. If you have more information for me to post here, please let me know, but I will not answer support questions.

Important! - Whenever you replace/update the BIOS for this board, you must reset the settings to default and reenter all of the BIOS tunings you have done! Just do the following:

1. Write down any settings you have changed from the default
2. Update the BIOS as normal according to the instructions with the flash program
3. Select the "Reset to default" option
4. Reprogram all of your settings

Other sites with general information about the Award 4.50PG BIOS, as used in this motherboard, include:

• Award's BIOS Error Messages
• Award's BIOS POST Codes
• Award's UMC 8881 Chipset Features
• Unicore's UMC 8881 Word document

ROM Types
It is important to use the correct type of BIOS upgrade because, allegedly, the Intel and SST Flash ROMs store the "ESCD" data (which is used to store plug and play data) at different addresses. It is therefore also important that jumper JP13 is set appropriately for your ROM type. If you flash with the wrong BIOS, it'll probably still work but your computer will have Plug and Play problems. Mine, for instance, kept forgetting the PS/2 mouse and serial ports existed and I would have to power off and unplug the mouse before it would re-detect them.

BIOS Versions

UUD951108

This was apparently the first UUD BIOS. Quantex has this version available at their web site.

• UUD951108i - Intel-Style update.
• UUD951108s - SST-Style update.
• FLASH.EXE - BIOS update utility.

UUD951205

I have no information about this version. Pionex Computers has this version available at their web site.

• UUD951205s - SST-Style update.

UUD951222

This version seems to have trouble with large (>1024 cylinder) hard disks.

UUD960104

I have no information about this version. Quantex has this version available at their web site.

• UUD960104i - Intel-Style update.
• UUD960104s - SST-Style update.
• FLASH.EXE - BIOS update utility.

UUD960111

I have no information on what features or fixes this version adds.

UUD960306

When I installed this BIOS version, Windows 95 re-detected my BIOS, hard disk controller, and system board, so I suspect it fixes/adds to these subsystems. Wintune 95 reported no appreciable differences in CPU speed before and after flashing, but the hard disk speed did increase by about 20%. Byte's ByteMark benchmark reported nearly identical CPU speeds before and after.

This version fixes the following problems:

• Support for an NCR SCSI card.
• Support for the memory parity check function.
• Support for NEC EDO DRAM.

Quantex has this version available at their web site.

• UUD960306i - Intel-Style update.
• FLASH.EXE - BIOS update utility.

UUD960326

When I flashed my BIOS to this version, it produced no noticable changes. Windows 95 did redetect my floppy controller, so it must have fixed some sort of floppy bug.

Only use this if U6 chip is marked "74LS244". Do not upgrade if U6 is marked "74F244".

This version fixes the following problems:

• Support for AMD 5x86-150MHz and 160MHz CPU. AMD never released these CPUs, but if you overclock your board to run the CPU ad 4x 40MHz, the BIOS reports it as an AMD 5x86-P90 at 160MHz!
• A fix for system boot-up problems when only 1MB of DRAM is installed on the board.

Quantex has this version available at their web site.

• UUD960326i - Intel-Style update.
• UUD960326s - SST-Style update.
• FLASH.EXE - BIOS update utility.

UUD960520

I am now using this version from the Biostar BBS. Note that my motherboard originally had the Intel-style BIOSes. I had to rejumper JP13 to the SST setting to get it to remember that my PS/2 mouse existed, though. I don't necessarily reccommend using the wrong BIOS type, but it seems to work.

Pantex has this version available at their web site.

• UUD960520s - SST-Style update.

• BIOS Features:

  Gate A20 Option - Fast

  A20 refers to the first 64KB of extended memory, also known as the high memory area. This option uses a faster way to access memory above 1 MB than the normal "keyboard controller" route.

  Set all others as appropriate

• Chipset Features:

  Auto Configuration - Disabled

  DRAM Read Wait States - 0

  My 80ns-rated RAM can take this. Lately, I've gone back to 1WS, though.

  DRAM Write Wait States - 0

  My 80ns-rated RAM can take this. Lately, I've gone back to 1WS, though.

  L2 Cache Wait States - 2-1-1-1

  Wait states and cache timings are horribly fun to try to figure out. I usually select the lowest possible and then turn them up if the PC crashes. In this case, my MB-8433UUD acts OK even with 0 wait states and 2-1-1-1 timing.

  Key Controller Clock - PCICLKI/4

  EDO DRAM Install Option - Disabled

  Enable only if you have EDO DRAM.

  L1 Cache Update Scheme - Wr-Back

  Early Cache Write Mode - Enabled

  System BIOS Cacheable - Enabled

  Video BIOS Cacheable - Enabled

  Host-to-PCI Post Write - 0WS

  Host-to-PCI Burst Write - Enabled

  PCI Bus Park Option - Enabled

  PCI Posted Memory Write - Enabled

  Slow Refresh (1/4 Freq) - Disabled

  Host Clock/PCI Clock - 1:1

  I overclocked my board to 40MHz and I had to move this down to 1 : 2/3 to get reliable performance. At that speed, my PCI bus is running at just 26.6MHz instead of the 33MHz it works at by default, but it's a fair trade for the performance gain of overclocking.

  Burst Copy-Back Option - Enabled

  Preempt PCI Master Option - Disabled

  IBC DEVSEL# Decode - Med

  Fast is faster but less stable. I've also heard reports that "Fast" will trash certain hard drives!

• The BIOS Survival Guide
• System Optimization-Information
• The BIOS Companion

CPU Types

According to the motherboard manual, the MB-8433UUD is capable of 25, 33, and 40 MHz system board speed, 1x, 2x, 3x, and 4x CPUs, and 25, 26.67, and 33 MHz PCI bus speed. It supports just about any 486-pinout CPU and is highly configurable using JP16, JP37, JP39, JP45, JP46, and RN11-RN15.

Cache

The MB-8433UUD has 9 sockets for SRAM cache. These can be configured with JP5, JP6, and JP7.

There has been some question on the net lately about so-called "fake cache" being installed on cheap motherboards. My MB-8433UUD has 9 socketed "real" asynchronous SRAM cache. My cache chips are labeled such:

UM61256FK-15
9547S
80503/M9T04

Memory

The MB-8433UUD has 4 72-pin SIMM sockets. It requires 80ns (70ns reccommended) SIMMs and supports EDO with the UUD960326i BIOS. I have a pair of unmatched 8MB SIMMs and they work fine.

Pantex Computer's Tech Base reports that TI memory doesn't work with the MB8433UUD.

BIOS/Chipset

Mine came with an Intel-style 12-volt flash ROM, but some may come with SST-style 5-volt flash ROM or just plain ROM instead. It uses the UMC 8881/8886 chipset. See the BIOS Versions section above for more information.

The ROM type is set with JP13.

There are apparently three versions of this board. One has a regular CMOS battery, the next has an integrated "Dallas" chip, and the third has some other sort of integrated CMOS chip. Can anyone be more specific about this?

Expansion

The board has 4 8MHz ISA slots and 3 PCI slots with room between them so they're not shared like on some motherboards.

Communications Ports

The board has UMC 8663 and 8667 chips which contain two 16550 UART serial ports and one SPP/EPP/ECP parallel port. These ports are just headers on the board and it comes with slot-cover ports with ribbon cables.

The parallel port's ECP DMA channel is configured with JP8 and JP9.

Pantex Computer's Tech Base reports that the Diamond SupraExpress 28.8 Modem's driver fails with this motherboard. However, I have reliable reports that people have had success with this modem, so this may not be true for all BIOS versions.

Keyboard/Mouse

It's got an AT keyboard and PS/2 mouse port as well. The keyboard port is in the right spot for a standard baby-AT motherboard. The PS/2 mouse port is right next to the keyboard port, though, so most cases can't get to it. There is also a header on the board for the PS/2 mouse and an optional ribbon cable to a slot-cover port.

I don't reccommend bothering to try to get the PS/2 mouse working. It has problems with the keyboard. Also, some people have reported trouble getting their older AT-style keyboards to work with this board. I am using a vintage 1984 IBM "battleship" keyboard with mine and have had no trouble. I have reports that some old AST and DTK keyboards failed. Your mileage may vary.

You can make your own mouse cable with the right parts. Here's the pinout of the PS/2 mouse port and the MB-8433UUD's PS/2 mouse port on the motherboard:

      PS/2 port                       MB-8433UUD Jumpers
(as the mouse sees it)                 Pin#  Connection
                                         1      Data
N/C----* # *----Clk                      2      N/C
VCC---*  #  *---Gnd                      3      Gnd
N/C-----* *-----Data                     4      VCC
                                         5      Clk

Storage

It's also got a built-in floppy controller and both primary and secondary Enhanced IDE controllers on the board. The chip on mine is marked as a UMC 8672 chip, but since that's a VESA controller, it's probably wrong. Others have reported a UMC 8673 instead. This is a single-fifo controller which limits performance somewhat. There are Windows 95, 3.1, OS/2, and NT drivers for this controller available and they work fine on my MB and seem to enhance performance somewhat.

According to the BIOS, this hard disk controller will do PIO modes 1 through 4, but I found this to not be the case. Windows 95 crashed whenever serious disk access occurred when I attached a disk capable of PIO mode 3 (a Quantum Fireball) but these crashes stopped when I set the BIOS to use PIO mode 2 instead. Maybe the new drivers can handle this speed?

Note for Linux, OS/2, and other real multitasking OS users: The MB-8433UUD does not use the buggy RZ1000 or CMD640B IDE controller chips! This means you don't have to worry about all the problems with these chips!

Jumpers

A complete list of jumper settings can be found on my Jumper Settings page.

The board is quite fast. Mine came with 256k of asynchronous SRAM cache (this is the so-called "real cache" operating in write-back mode, not the fake stuff some vendors are selling as "write-through cache") and performs quite well compared to other boards. At 133 MHz, my 5x86 performs like a (fictitious) 80 MHz Pentium in integer operations, though its floating-point performance is far lower. This is to be expected from a 486 core with Writeback cache. To find out more info about your AMD 5x86 CPU, try 5x86info.zip.

My 5x86 chip is an ADW chip, not the cooler (?) ADZ variety, but it runs plenty cool. Even without the CPU fan attached it is cool enough to touch. Don't try that with a Pentium! I do run it with a CPU fan, though, just to keep it extra cool.

This puts it in the Pentium 90 area for integer performance. In fact, with the 5x86 overclocked to 160, the Award BIOS reports the CPU at bootup as an AMD 5x86-P90 at 160MHz!

I had to take the PCI bus speed down in order to get reliable boots, though. The next-highest multiplier for the Host Clock/PCI Clock setting in the Chipset Features menu of the BIOS is "1 : 2/3". This yields a PCI bus speed of just 26.6MHz, but I think the trade-off was worth it!

I also had to add 1 wait-state for both reading and writing to DRAM and move the L2 cache timing to 3-2-2-2 instead of 2-1-1-1. Again, this slows the board down somewhat, but the added CPU cycles more than make up for the gap.

PS/2 Mouse and Keyboard Hang

Under Windows 95, the keyboard and PS/2 mouse conflict causing a complete loss of control. If you press a key, a mouse button, and move the mouse at the same time, both the keyboard and mouse become unusable. If you then warm-boot the machine, it will "update ESCD" and remove PS/2 mouse support! You have to cold-boot to get the mouse back! And, obviously, without mouse or keyboard, you can't shut down Windows nicely. This is, in a word, bad. I could repeat this problem in Windows 95's real mode (DOS prompt), under DOS 6.22, and in Linux under XFree86, so it's not just a Windows 95 driver issue.

Quantex reports this problem as unresolved on their page of support info. They say it's a bug in the UMC I/O controllers, and I believe it.

This problem can be worked around by simply using a serial mouse instead of a PS/2 mouse. I'm running fine with an old Logitech mouse on COM2 right now.

No PS/2 port on the board

The mouse port is located physically right next to the keyboard port. If your case doesn't have a cutout there for it (and most don't), you have to get an optional cable which moves the port to an expansion slot cover. This cable costs extra from the manufacturer so the company you buy the board from will probably charge you extra, too. I have the cable (they threw it in when they replaced my first board) and the pinout is in the manual (and above) if you want to make your own.

EIDE driver conflicts with Novell Client32

Another problem with this board is that the EIDE driver (um8673.sys) conflicts with the Novell Client32 network drivers. Novell users must use the Client32 drivers with this board because of the lack of upper memory (640 - 1MB) space. The standard Novell VLMs will use up all upper memory and leave a large portion of them in low memory, leaving very little RAM. The Client32 drivers load into extended memory, eliminating this problem.

Windows 95 Hangs/Crashes on Shutdown

CyberMax Computers' Top Ten Tech Support Answers says that Windows 95 will hang on shutdown and reboot if you set the Video BIOS Cacheable and Video BIOS Shadow to "on" in the BIOS setup. They claim this happens with the AMD DX4-100, DX4-120, and 5x86-133 CPUs. However, I have a 5x86-133 and I have never had this problem.

Please don't write to me asking for tech support. I don't work for Biostar, and everything I know about them and their board is right here! Please do send me mail if you have any information to contribute, though!

sfoskett@slf.gweep.net Back to my hardware page Back to my home page