🚀 go-pugleaf

David Longley <David@longley.demon.co.uk> wrote:

>PCI 1 - - Miro DC20 (IRQ 9)

>Audio - - IRQ 5

You forgot IRQ 2, which is used by the MIDI port, and just happens to be
the same as IRQ 9...

possibly.

-- 
Marcus Bainbridge 
marcus@mwb.u-net.com - http://www.u-net.com/~mwb

Does anyone know if the Triton II chip sets (430HX PCIset and 430VX
chip set) cache more than the "Lower 64MB" of memory, which is what the
Triton (430FX PCIset) caches.

I found out about this when I found my Pentium 166MHz system
ran slower with 128MBs than with 64MBs, 22 versus 37 on Norton
Utility's System Information benchmark.  After having the retailer
check all of the EPROM settings, we called the motherboard vendor
(Tyan for Titan III motherboard) and found out that the Triton
chip set only cached "the first" 64MB of RAM.

I found a web site that had the complete documentation on-line
(which www.intel.com doesn't seem to have: its down now so I
 can't check again.)

Page 22 of the documentation revealed that only the "lower" 64MB
of memory was cached.  There was no earlier direct reference to the
fact (I suppose that counting bits in registers would have made it
clear), although there were many references to supporting caching
of various types and 128MB of memory.

I can't find the Triton II complete spec sheet, so does anyone know
how much memory is elegible for caching in  430HX and 430VX PCIset's?

Thanks.

(I but "lower" and "first" in quotes above since I don't know where
 the memory for the Norton benchmenk is allocated.  I would have
 assumed that the memory would be within the first 64MB of memory on
 a system with nothing but Windows 95 and the Norton test running, but
 it might work just the opposite.  Having more than 64MB is not
 supposed to turn off caching - it still should remain active for some
 64MB of memory, so if the memory allocation policy was the opposite
 to what it is, the benchmark should have shown no difference between
 performance with 64MB and 128MB.)

> Does anyone know if the Triton II chip sets (430HX PCIset and 430VX
> chip set) cache more than the "Lower 64MB" of memory, which is what the
> Triton (430FX PCIset) caches.


430hx supports 11 bits of tag cache, and that's not counting status,
which it implements internally.  the top three bits are described 
as optional (and are omitted from the 430vx.)  since you can get 
512k cache for a pittance these days, they both support 128M
cachable range.  I don't see specs at this level of detail for 
the 430fx, but I suspect it also implements 8 bits, providing
128M cachable with 512k cache.

regards, mark hahn.
--
operator may differ from spokesperson.	hahn@neurocog.lrdc.pitt.edu
					http://neurocog.lrdc.pitt.edu/~hahn/

In article <833406694snz@longley.demon.co.uk>
           David@longley.demon.co.uk "David Longley" writes:

> I have had frequent problems with plug and play IRQ etc especially with
> PCI slots on my Endeavour Board. It's not just this one either, it's been
> a feature of all 4 I have set up.
> 
> Often it's just the video card. Switch on and no output. Sometimes this
> clears by setting DIP 4 to on to clear NVRAM, but not always. 
> 
> Today, I was adding a Miro DC20 PCI card. 
> 
> System
> 
> ISA 1 - - Digital Turbo 3 network card (IRQ 10)
> ISA 2 - - empty
> ISA 3 - - empty
> PCI 1 - - Miro DC20 (IRQ 9)
> PCI 2 - - Matrox Millenium Video
> PCI 3 - - empty
> PCI 4 - - Adaptec AHA 2940 SCSI (IRQ 11) 
> 
> Audio - - IRQ 5
> 
> Problems started with the Miro grabbing INT 9. I was a bit concerned and
> rebooted to enter setup. I set IRQ 9 to be available to ISA. The Miro
> still grabbed it and I then lost the Miro driver when I went into
> PREMIERE LE.
> 
> Now the system won't boot at all. I have pulled out the RAM, pulled all
> the cards, power leads and them mains lead out. Re-assembling it, I do
> not get a signal out to the monitor.
> 
> I've even taken the battery out hoping that will clear whatever is in
> whatever RAM is on the board.
> 
> Can anyone explain why this happens? (and it has happened to others too).
> 
> Can anyone suggest what to do when this does happen as it's pretty soul
> destroying.
> 
> 
> 
I have had to strip the system down to the board, 32MB EDO, 256K COAST
a speaker, floppy drive and P166.

I am trying to reflash the BIOS.

The motherboard is a socket 7 PBA641508-701 , CA639903-001.

With J1N1 jumpered 1-3, 5-6 I get nothing. I have jumpered 2-3 for 
Flash recovery, and I have made a bootable floppt. I have also unpacked
to that floppy. the files from the latest BIOS EXE (10005cb0.exe).

When I boot, I get one bleep then another some time after, and then 
a series of regular low pitch beeps, wshich, from plugging the disk into
another machine is which does not support flash upgrade, is indicating that
the system does not support flash upgrading!!!

Help.....Please....
-- 
David LOngley

In article <833406694snz@longley.demon.co.uk>
           David@longley.demon.co.uk "David Longley" writes:

> I have had frequent problems with plug and play IRQ etc especially with
> PCI slots on my Endeavour Board. It's not just this one either, it's been
> a feature of all 4 I have set up.
> 
> Often it's just the video card. Switch on and no output. Sometimes this
> clears by setting DIP 4 to on to clear NVRAM, but not always. 
> 
> Today, I was adding a Miro DC20 PCI card. 
> 
> System
> 
> ISA 1 - - Digital Turbo 3 network card (IRQ 10)
> ISA 2 - - empty
> ISA 3 - - empty
> PCI 1 - - Miro DC20 (IRQ 9)
> PCI 2 - - Matrox Millenium Video
> PCI 3 - - empty
> PCI 4 - - Adaptec AHA 2940 SCSI (IRQ 11) 
> 
> Audio - - IRQ 5
> 
> Problems started with the Miro grabbing INT 9. I was a bit concerned and
> rebooted to enter setup. I set IRQ 9 to be available to ISA. The Miro
> still grabbed it and I then lost the Miro driver when I went into
> PREMIERE LE.
> 
> Now the system won't boot at all. I have pulled out the RAM, pulled all
> the cards, power leads and them mains lead out. Re-assembling it, I do
> not get a signal out to the monitor.
> 
> I've even taken the battery out hoping that will clear whatever is in
> whatever RAM is on the board.
> 
> Can anyone explain why this happens? (and it has happened to others too).
> 
> Can anyone suggest what to do when this does happen as it's pretty soul
> destroying.
> 
> 
> 
My copy of the latest BIOS upgrade was corrupt!!. All was restored to
normal after I used the previous BIOS upgrade. It is a pain that the
BIOS can get corrupted so easily on these boards though....

Is the version 5 corrupted at intel.com?
-- 
David LOngley