Just a quick update..

Got the Asus PCI/I-486SP3G loaded up with RAM - all slots in use for a total of 32MB. After removing the sticker on the BIOS chip, I don't see a UV window so I'll try to flash it to the latest 1999 BIOS this evening. If it doesn't work, I'll need to burn a new a new chip and go from there. I did throw an ATI Rage PCI card in there, my CF->IDE card (directly connected to the IDE controller native to the motherboard), and the 100Mhx AMD processor and all appears well (despite the BIOS not being updated to allow write-back mode).

Going back to the ASUS VL/I-486SV2GX4 side of things, a VOGONS member has graciously burned some new EPROM chips for me and those should be arriving soon! Once the new chips are dropped in place on these boards I'll continue the benchmarking - most likely using my Intel DX4-100 in write-back mode vs the AMD alternative.

So... After trying PFlash, AWDFlash and UniFlash, none of these utilities recognize the flash type of the existing chip so the PCI/I-486SP3G cannot be updated this way. Thus, a third chip is being burned for me with the latest BIOS. When these chips roll in I should have everything sorted out (crossing fingers) and ready to benchmark! Stay thirsty my friends... 😀

Artex wrote:

So... After trying PFlash, AWDFlash and UniFlash, none of these utilities recognize the flash type of the existing chip so the PCI/I-486SP3G cannot be updated this way. Thus, a third chip is being burned for me with the latest BIOS. When these chips roll in I should have everything sorted out (crossing fingers) and ready to benchmark! Stay thirsty my friends... 😀

Older boards like this you need to use an old version of Aflash.exe and you need to move the jumper on your board to allow programming (flashing). This is from my experience with flashing Socket 5 boards from ASUS. Since this board is from the same time period from the same manufacturer I would expect it to be similar.

Also by looking at the BIOS releases of this board I don't think it's really needed unless you want to get it Y2K ready.

EDIT: By looking the schematics and pictures there don't seem to be any jumper which you need to move to allow flashing. I'll send you a 1998 version of aflash.exe which works with my Socket 5 boards from ASUS when I get home from work.

vetz wrote:

Older boards like this you need to use an old version of Aflash.exe and you need to move the jumper on your board to allow progr […]
Show full quote

Artex wrote:

So... After trying PFlash, AWDFlash and UniFlash, none of these utilities recognize the flash type of the existing chip so the PCI/I-486SP3G cannot be updated this way. Thus, a third chip is being burned for me with the latest BIOS. When these chips roll in I should have everything sorted out (crossing fingers) and ready to benchmark! Stay thirsty my friends... 😀

Older boards like this you need to use an old version of Aflash.exe and you need to move the jumper on your board to allow programming (flashing). This is from my experience with flashing Socket 5 boards from ASUS. Since this board is from the same time period from the same manufacturer I would expect it to be similar.

Also by looking at the BIOS releases of this board I don't think it's really needed unless you want to get it Y2K ready.

EDIT: By looking the schematics and pictures there don't seem to be any jumper which you need to move to allow flashing. I'll send you a 1998 version of aflash.exe which works with my Socket 5 boards from ASUS when I get home from work.

Yeah, I'd like to bring it to the latest version if I can. I used a AWDFlash version from 1999.

Try this version 😀

Prime example of... "I should have RTFM." 😀 Gonna try this shortly with your version of AFlash. There IS a jumper.

So you have the MANUAL?? You better scan & upload that to the internet, the one on ASUS.com sucks and is missing loads of pages, including the ones you just posted.

You can't use my version of AFLASH.EXE. You need flash.exe, a previous version. I've attached it.

Btw, I thought that jumper was just for selection which type of BIOS chip you used, not for actual flashing. The Socket 5 boards have this jumper AND a programming jumper.

Totally forgot I had the manual (I received this board last year). It's a pretty large manual so it will take some time, but I agree its probably worth it. I also have the manual for the ASUS VL/I-486SV2GX4. I'll try the Flash utility (why are there so many of these?!!) and let you know the result. I think the jumper is required (according to the manual) and acts as a fail-safe: Right now it's set to +5V and all these utils can't recognize the flash type. I bet if I change it to 12V this flash will work.

Success!! BIOS has been updated successfully after moving that jumper to 12V and using the Flash utility. Still not running the Intel DX4 (&EW) or the AMD 100Mhz processor in Write-back mode though - even though the BIOS is properly configured for L1 Write-back.

Artex wrote:

Success!! BIOS has been updated successfully after moving that jumper to 12V and using the Flash utility. Still not running the Intel DX4 (&EW) or the AMD 100Mhz processor in Write-back mode though - even though the BIOS is properly configured for L1 Write-back.

Great! 😎

You've double checked the jumpers right?

EDIT:
According to Google Groups the Saturn chipset does not support L1 WB:

Also, the DX4/133 benifits greatly from installing a dirty tag ram on the
L2 cache and running the L2 cache in Write Back. The CPU will not run
in WB (ie. L1 cache), as the Saturn chipset does not support it

Source

Several posts mentions the same thing.

Btw, I would LOVE to see a full ASUS VL/I-486SV2GX4 manual online. Even the pictures you took of the two pages above is miles beyond what ASUS have on their website.

Ok, I will start scanning both manuals. A shame that the board can't do L1 WB - I wonder how much of performance hit that will be.

Artex wrote:

Ok, I will start scanning both manuals. A shame that the board can't do L1 WB - I wonder how much of performance hit that will be.

I*m searching atm if it's true or not, but here are the jumpers needed for unofficial AMD-X5-133 support:

JUMPER SETTING OF PCI/I-486SP3G FOR AMD-X5-133 ============================================== […]
Show full quote

JUMPER SETTING OF PCI/I-486SP3G FOR AMD-X5-133
==============================================

- CPU external clock: 33MHz
JP18 1&2
JP20 1&2

- It appears that non-SL setting is necessary for soft-reset to work
on
Linux and OS/2.

CPU SL selector: non-SL
JP30 2&3
short JP31:pin1 with JP32:pin1

- A x2 setting will tell the AMD-X5-133 to quadruple

CPU internal clock selector: x2
JP33 2&3

v1.6 MOTHERBOARD SPECIFIC v1.8 MOTHERBOARD SPECIFIC
========================= =========================

Copy code to clipboard

1                                       - voltage selector: 3.45v
2                                         JP40 2&3                
3                                                         
4- CPU type selector:                   - CPU type selector:     
5- JP34 2&3                               JP34 2&3                
6- JP35 1&2                               JP35 1&2                
7- JP36 connect pin1&3 with               JP36 2&3       
8  a 1kohm resistor    
9- JP38 open                            - JP38 open

Mode of L1 cache
----------------

JP36 controls the L1 cache mode of the AMD-X5-133. It has to be pulled
low to put the CPU in write-through mode. It appears that ASUS has got
it right with v1.8, but not on v1.6. Many thanks to Jochen Kiemes
<10011...@compuserve.com> who actually took some voltage
measurements and suggested the 1kohm resistor.

The setting of the cache in the CMOS BIOS refers to the L2 cache,
which can be in write-back mode. However, the L1 cache must be in
write-through mode. You can tell which mode your AMD-X5-133 is in by:

1) MSDOS: using 5x86info posted in alt.comp.periphs.mainboard.asus or
shareware utility AGSI
(e.g.
ftp://ftp.informatik.hu-berlin.de/pub/pc/msdo … ls/agsi114e.zip)

2) Linux: "cat /proc/cpuinfo". The reported model for different L1
cache modes are:
Linux version
1.2.13 1.3.x x<99 1.3.x x>99
write-through 4 14 Am5x86-WT
write-back 5 15 Am5x86-WB

ISA CAT Mode
------------

Finally, disable ISA GAT Mode in CMOS setup. Some people reported
problems with sound card otherwise.

Artex wrote:

I wonder how much of performance hit that will be.

According to an article in C'to they confirm the 486SP3G does not support L1 WB. Performance benchmark indicates a 8-10% performance loss. Read the whole bench here:
https://groups.google.com/d/msg/alt.comp.peri … Uc/_7jzL5BresYJ

Interesting findings today Vetz... I appreciate all of your help (again!). I did see the article about the jumper config for the AMD-X5-133 and I think I'll end up using this processor to max out the board. Throws a little wrench into the benchmarks when comparing the Asus PCI/I-486SP3G against the Biostar MB-8433UUD now that we know the Asus board won't do L1 write-back. I wonder if the memory interleaving will help make up for it?

Asus PCI/I-486SP3G
AMD X5-133
Write Through L1 mode
Interleaving enabled (all slots filled)

Asus PCI/I-486SP3G
AMD X5-133
Write Through L1 mode
Interleaving disabled

Biostar MB-8433UUD
AMD X5-133
Write Through L1 mode

Biostar MB-8433UUD
AMD X5-133
Write Back L1 mode

vetz wrote:

Artex wrote:

I wonder how much of performance hit that will be.

According to an article in C'to they confirm the 486SP3G does not support L1 WB. Performance benchmark indicates a 8-10% performance loss. Read the whole bench here:
https://groups.google.com/d/msg/alt.comp.peri … Uc/_7jzL5BresYJ

So from that post, it looks like the AMD-X5-133 and Intel DX4-100(&EW) won't work with L1 WB, but the AMD486DX4-100SV8B (8KB WB) will?

Artex wrote:

So from that post, it looks like the AMD-X5-133 and Intel DX4-100(&EW) won't work with L1 WB, but the AMD486DX4-100SV8B (8KB WB) will?

Hmm.. looks like it. This page does not mention the SV8B when talking about WT mode being required (only the EW and P24T). The SV8B uses jumper J36 differently than the Intel CPU's. Could be that WB works with AMD DX4/100 and not Intel. I suggest you doublecheck your jumpers and give it a try.
http://www.elhvb.com/mobokive/archive/asus/48 … g/sp3g-cpu.html

vetz wrote:

Artex wrote:

So from that post, it looks like the AMD-X5-133 and Intel DX4-100(&EW) won't work with L1 WB, but the AMD486DX4-100SV8B (8KB WB) will?

Hmm.. looks like it. This page does not mention the SV8B when talking about WT mode being required (only the EW and P24T). The SV8B uses jumper J36 differently than the Intel CPU's. Could be that WB works with AMD DX4/100 and not Intel. I suggest you doublecheck your jumpers and give it a try.
http://www.elhvb.com/mobokive/archive/asus/48 … g/sp3g-cpu.html

Interesting, so the AMD 100SV8B DOES work in WB mode, but what's odd is that although the board is jumpered for 3.0 x 33, the bus speed is only 31.6Mhz (31.6 x 3.0 = 94.8Mhz).

Really strange! Does it boot up as 100mhz during POST?

Yeah, definitely shows it at 100Mhz at post. Odd eh?

Found this info today on the SP3G:

Q04) What is the better cache sheme: write-through or write-back ?
How can I upgrade my SP3g with a dirty Tag RAM ? Does it increase
the performance ?

The ASUS SP3G can drive write-through or write-back cache. You can use both
modes with and without dirty tag RAM present. In write-through mode, dirty
tag RAM has no function. The BIOS default is "write-back", resulting in
some 5% performance loss, because if there is no dirty tag RAM, then there
are lots of unnecessary memory writes, since each cache line refill will
need to be preceeded by a write back of four DWORDs, even if none of them
had been modified.
In write-back mode, dirty tag RAM increases real-world performance by
approx. 1% over write-through. The small positive effect of write-back on
PC compatible hardware is caused by the lack of write-allocate. In case
of the Saturn chipset, the number of clocks for some transactions is
higher for write-back and the Saturn has already very efficient write
posting buffers which compensate much of the inefficiency of not having a
dirty tag RAM.
Starting with rev. 1.8, the SP3G got a socket for a dirty tag RAM. With
this RAM installed. Suitable chips for 256kB cache size are 64K x 1 RAMs
with 15 ns access time, e.g. Cypress Cy7C187, AT&T ATT7C187-15,
Matra MHS HME 65787-15, Motorola MCM6287C-15 and Toshiba TC5562-15 or any
other equivalent.

-----------

Q05) What type of simms can the PCI/I-SP3g use ?

The SP3g can only use pairs of simm modules, because the SATURM II chipset
alsways enables memory interleave, and to do so you need at least to
modules. So if you plan for example 16MB for your system count the price
of 2 * 8MB.
The board can use simms with or without parity as it doesn't use the parity
bit on the simms.
With 4MB and 16MB modules you can either use pairs of single sides or
double sided ( = density, I like that one Stefan). Even though people
reported that the successfully mixed a single sided and double sided 16MB
module on one memorybank.
With 8MB and 32MB modules you can only use double sided (= density) simms.