440 chipset and hard drive limitation

Discussion in 'Gigabyte' started by Mark M, Apr 4, 2004.

  1. Mark M

    Mark M Guest

    So can you explain why Gigabyte are saying there in a 75 GB limit
    on this board which their BIOS upgrade overcomes?
    Mark M, Apr 13, 2004
    1. Advertisements

  2. Because some clown sent them the revised bios saying that it
    resolved the 75GB limitation (or something similar obscure).
    Folkert Rienstra, Apr 15, 2004
    1. Advertisements

  3. Mark M

    Mark M Guest

    Well Folkert you wont believe this but the last email I got from
    Gigabyte tech support said this:

    I am correct i have one of these boards myself.

    Question :
    I think you are wrong when you say "im afraid the 440 chipset only
    allows up to 75GB".

    Maybe this belief pervades the whole of the Gigabyte organization?
    Maybe, just maybe, they are correct.

    But I am left wondering what the hell they are talking about and
    why they say there is a 75 GB limit. Any ideas what they could be
    referring to (bearng in mind that a BIOS flash is said to fix it)?

    Surely it could not be that they are unable to write "137 GB" or
    "128 GB" and instead they wrote 75 GB by mistake?
    Mark M, Apr 15, 2004
  4. Mark M

    Mr. Grinch Guest

    I've been following this thread and I don't know what's going on at
    Gigabyte but either a) they really screwed up when they wrote the bios or
    b) the tech response you received is wrong.

    I've got a Tyan Tiger 100 S1832DL motherboard, thats a 440BX Dual. It has
    the ATA drive size limit you would expect at 128 / 137 GB, depending on
    your method of measuring HD size. It does NOT have a limit at 75GB. I've
    verified this by installing 160GB drives on the board controller and
    comming up against the 128 / 137 limit.

    From the Maxtor site:

    So I don't know about any BX boards with a 75GB limit, but I can say for
    certain my BX board doesn't.
    Mr. Grinch, Apr 15, 2004
  5. Aha, so Gigabyte tech support is a one person operation.
    And now that person is covering for himself.
    We already established that they weren't. It's not the chipset.
    Read my response to Roger.
    Folkert Rienstra, Apr 15, 2004
  6. Mark M

    Mark M Guest

    -- snip --

    I must confess I don't really understand the details of what you
    have written to him. To take a guess maybe "P-CHS" means "physical
    CHS" and you are referring to an emulation of the physcial CHS.

    Anyway, it seems it may lie deep in the BIOS so I will just accept
    that the BIOS flash fixes it. I can't try the BIOS flash for a
    while so I will have to wait to see if it really does do the trick.
    Mark M, Apr 15, 2004
  7. L-CHS is the CHS as used by the software interface (int13 etc.)
    P-CHS is a translated CHS as used by the harddrive interface.

    For some odd reason software makers and hardware makers went a dif-
    ferent route expanding the old 20-bit (528MB) CHS (1024 16 63) to 24-
    bit (8GB) CHS. L-CHS expanded the Heads register with 4 bits (1024 256
    63) and P-CHS expanded the Cylinder register with 6 bits (65536 16 63).
    Although 2 bits more P-CHS is still limited to C*H*S= 8GB.
    Folkert Rienstra, Apr 16, 2004
  8. Both Folkert, and I think that this problem, is probably down to a numeric
    'translation' limit inside the BIOS. There is definately no 'chipset' limit
    at this point (given that a BIOS change fixes it, and thousands of other
    boards with the same chipset do not have a problem). Basically, there are a
    number of different 'values' used inside the IDE code. There are the
    'numbers' used to refer to cylinders, heads etc., that are translated inside
    the drive to the 'real' geometry of the drive, and are also translated from
    'LBA' values inside the BIOS. Now (for instance, not a real example...),
    suppose you elected to take the values at one point in the BIOS, and do the
    arithmetic using BCD, and use a register at some point in the maths, that
    limited the maximum value in the 'maths' to 99999999. Since you are dealing
    with 'sector counts', this would give a drive size 'limit' of 50GB. This
    would be entirely a 'BIOS' problem (trying to blame it on the chipset, would
    just be down to wanting to find somebody else to blame for the programmers
    error). 75G, is a very 'odd' number in these terms, since generally all the
    arithmetic is binary (with some silly 'oddities' in the actual numbers
    allowed, which are the result of 'history' in the interface, and the upgrade
    methods chosen at times). It sounds as though perhaps one part of the
    conversion code, is thorougly 'screwed'...

    Best Wishes
    Roger Hamlett, Apr 16, 2004
  9. Btw: doing what trick?
    You "have not come across a 75 GB limit" so what do you think you'll notice?
    Folkert Rienstra, Apr 16, 2004
  10. Mark M

    Mark M Guest

    I hope that flashing this mobo's BIOS will allow it to see the
    space about 75 GB on the drive.

    After I flashed it I still can't see all of a 160 GB drive.

    Maybe it has a 128/137 GB limit now.
    Mark M, Apr 17, 2004
  11. Mark M

    Eric Gisin Guest

    You have an ancient mainboard. Why do you expect more than 135GB?

    Maybe you should follow all the other threads about LBA-48 and BIOS support.
    Eric Gisin, Apr 17, 2004
  12. Mark M

    Mark M Guest

    You are right. The mobo is far too old. I think it may destined
    for the trash sometime soon.

    Although it has no cash value, I had originally wanted to use the
    mobo in a machine to get on the net when my main machine was down.
    Mark M, Apr 17, 2004
  13. Why not just invest in a cheap ATA100/133 PCI card then? A second-hand
    one in particular would sell for next to nothing. Both promise and
    highpoint have issued BIOS updates which add 48bit LBA support for their
    range of products AFAIK.

    Nikolaos Tampakis, Apr 17, 2004
    Folkert Rienstra, Apr 19, 2004
  15. Mark M

    Regal Guest

    Are there any new mobos around nowadays for Intel's "Slot One" cpu?
    Regal, Apr 22, 2004
  16. New in condition maybe, from leftover stock, but I doubt they've been
    produced in years. AFAIK intel's latest were i820/i840 boards and on the
    VIA side it was boards with the Apollo Pro 133A, and this was like 3,5
    years ago.

    Nikolaos Tampakis, Apr 22, 2004
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.