• GALAXEUS: New DOS / Bare-Metal Arcade Engine gives solid 78Hz/78 FPS on a 386(sx) CPU (Hi-Res 544x480 on slow 8-Bit ISA Std. VGA)

    From Louis Ohland@ohland@charter.net to comp.sys.ibm.ps2.hardware on Thu Jul 30 06:58:13 2026
  • From Louis Ohland@ohland@charter.net to comp.sys.ibm.ps2.hardware on Thu Jul 30 07:05:22 2026
    From Newsgroup: comp.sys.ibm.ps2.hardware

    Would this be something significant for the 9533?

    Louis Ohland wrote:
    https://forum.vcfed.org/index.php?threads/galaxeus-new-dos-bare-metal-arcade-engine-gives-solid-78hz-78-fps-on-a-386-sx-cpu-hi-res-544x480-on-slow-8-bit-isa-std-vga.1258384/

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Josh Rodd@joshua.at.rodd.dot.us@rodd.us to comp.sys.ibm.ps2.hardware on Thu Jul 30 18:08:58 2026
    From Newsgroup: comp.sys.ibm.ps2.hardware

    Louis Ohland <ohland@charter.net> wrote:
    Would this be something significant for the 9533?

    Louis Ohland wrote:
    https://forum.vcfed.org/index.php?threads/galaxeus-new-dos-bare-metal-arcade-engine-gives-solid-78hz-78-fps-on-a-386-sx-cpu-hi-res-544x480-on-slow-8-bit-isa-std-vga.1258384/


    Well, the 9533 has a quite capable CPU in it (a 50Mhz Blue Lightning)
    using 60ns or 70ns FPM RAM (two cycles of the 25MHz external bus) on a not-so-capable ISA 8MHz bus.

    However, the XGA(-2) on the PS/2 E is attached some "funny" way, because
    it certainly performed a lot better than a typical ISA Super VGA would.
    Maybe that was just the XGA acceleration fooling me, but it had no
    problems at all keeping up with software rendered M-Motion in 16-bit
    direct colour.

    With that said, this project probably isn't too relevant to the PS/2 E
    since it managed to be a lot faster than a typical 386SX and in
    particular has a much more capable video system. It's actually a bit of
    a genius design - the only thing slow are peripherals that are dog-slow
    anyway, like the IBM 54G0135 drive, spinning just over 3,600 RPM and
    running at about 1MB/sec.

    Stacked up next to an MCA bus ThinkPad 720, the PS/2 E's video system
    performed substantially better than the 720's MCA-attached WD video
    chip... the same cannot be said for the ESDI drive in the ThinkPad 720, probably because that drive could actually do DMA.

    Incidentally the PS/2 E was a tragedy when they were being liquidated at
    Best Buy... 4MB machines with 1MB spent on the video subsystem, booting
    OS/2, and just swapping like crazy to a laptop-speed drive doing
    programmed I/O. A 71F7010 4MB upgrade made the thing actually usable.
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Josh Rodd@joshua.at.rodd.dot.us@rodd.us to comp.sys.ibm.ps2.hardware on Thu Jul 30 18:09:12 2026
    From Newsgroup: comp.sys.ibm.ps2.hardware

    Louis Ohland <ohland@charter.net> wrote:
    Would this be something significant for the 9533?

    Louis Ohland wrote:
    https://forum.vcfed.org/index.php?threads/galaxeus-new-dos-bare-metal-arcade-engine-gives-solid-78hz-78-fps-on-a-386-sx-cpu-hi-res-544x480-on-slow-8-bit-isa-std-vga.1258384/


    Well, the 9533 has a quite capable CPU in it (a 50Mhz Blue Lightning)
    using 60ns or 70ns FPM RAM (two cycles of the 25MHz external bus) on a not-so-capable ISA 8MHz bus.

    However, the XGA(-2) on the PS/2 E is attached some "funny" way, because
    it certainly performed a lot better than a typical ISA Super VGA would.
    Maybe that was just the XGA acceleration fooling me, but it had no
    problems at all keeping up with software rendered M-Motion in 16-bit
    direct colour.

    With that said, this project probably isn't too relevant to the PS/2 E
    since it managed to be a lot faster than a typical 386SX and in
    particular has a much more capable video system. It's actually a bit of
    a genius design - the only thing slow are peripherals that are dog-slow
    anyway, like the IBM 54G0135 drive, spinning just over 3,600 RPM and
    running at about 1MB/sec.

    Stacked up next to an MCA bus ThinkPad 720, the PS/2 E's video system
    performed substantially better than the 720's MCA-attached WD video
    chip... the same cannot be said for the ESDI drive in the ThinkPad 720, probably because that drive could actually do DMA.

    Incidentally the PS/2 E was a tragedy when they were being liquidated at
    Best Buy... 4MB machines with 1MB spent on the video subsystem, booting
    OS/2, and just swapping like crazy to a laptop-speed drive doing
    programmed I/O. A 71F7010 4MB upgrade made the thing actually usable.
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Kevin Bowling@kevin.bowling@kev009.com to comp.sys.ibm.ps2.hardware on Thu Jul 30 15:42:09 2026
    From Newsgroup: comp.sys.ibm.ps2.hardware

    On 7/30/26 11:09, Josh Rodd wrote:
    Louis Ohland <ohland@charter.net> wrote:
    Would this be something significant for the 9533?

    Louis Ohland wrote:
    https://forum.vcfed.org/index.php?threads/galaxeus-new-dos-bare-metal-arcade-engine-gives-solid-78hz-78-fps-on-a-386-sx-cpu-hi-res-544x480-on-slow-8-bit-isa-std-vga.1258384/


    Well, the 9533 has a quite capable CPU in it (a 50Mhz Blue Lightning)
    using 60ns or 70ns FPM RAM (two cycles of the 25MHz external bus) on a not-so-capable ISA 8MHz bus.

    However, the XGA(-2) on the PS/2 E is attached some "funny" way, because
    it certainly performed a lot better than a typical ISA Super VGA would.
    Maybe that was just the XGA acceleration fooling me, but it had no
    problems at all keeping up with software rendered M-Motion in 16-bit
    direct colour.

    Could it be directly CPU bus attached? I would guess it would show up
    in a particular way if it was ISA attached.

    With that said, this project probably isn't too relevant to the PS/2 E
    since it managed to be a lot faster than a typical 386SX and in
    particular has a much more capable video system. It's actually a bit of
    a genius design - the only thing slow are peripherals that are dog-slow anyway, like the IBM 54G0135 drive, spinning just over 3,600 RPM and
    running at about 1MB/sec.

    Stacked up next to an MCA bus ThinkPad 720, the PS/2 E's video system performed substantially better than the 720's MCA-attached WD video
    chip... the same cannot be said for the ESDI drive in the ThinkPad 720, probably because that drive could actually do DMA.

    Incidentally the PS/2 E was a tragedy when they were being liquidated at
    Best Buy... 4MB machines with 1MB spent on the video subsystem, booting
    OS/2, and just swapping like crazy to a laptop-speed drive doing
    programmed I/O. A 71F7010 4MB upgrade made the thing actually usable.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Josh Rodd@joshua.at.rodd.dot.us@rodd.us to comp.sys.ibm.ps2.hardware on Sun Aug 2 23:55:14 2026
    From Newsgroup: comp.sys.ibm.ps2.hardware

    Kevin Bowling <kevin.bowling@kev009.com> wrote:
    Could it be directly CPU bus attached? I would guess it would show up
    in a particular way if it was ISA attached.

    I have guessed that it was, but I haven't touched a PS/2 E in the flesh
    since the late 1990s.

    "Spiritually" I suspect it derives a lot from the Model 35 SLC design,
    which had a 84F7985 IBM SVGA which looks to be attached to a local bus.
    The PS/2 E was positioned to be "premium", hence why it had the faster
    XGA-2 versus the cheap SVGA chipset.

    (Didn't some Mod. 30s and Mod. 45s use the same planar as the PS/2 E?)

    With that said, this project probably isn't too relevant to the PS/2 E since it managed to be a lot faster than a typical 386SX and in
    particular has a much more capable video system. It's actually a bit of
    a genius design - the only thing slow are peripherals that are dog-slow anyway, like the IBM 54G0135 drive, spinning just over 3,600 RPM and running at about 1MB/sec.

    Stacked up next to an MCA bus ThinkPad 720, the PS/2 E's video system performed substantially better than the 720's MCA-attached WD video
    chip... the same cannot be said for the ESDI drive in the ThinkPad 720, probably because that drive could actually do DMA.

    Incidentally the PS/2 E was a tragedy when they were being liquidated at Best Buy... 4MB machines with 1MB spent on the video subsystem, booting OS/2, and just swapping like crazy to a laptop-speed drive doing
    programmed I/O. A 71F7010 4MB upgrade made the thing actually usable.


    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Louis Ohland@ohland@charter.net to comp.sys.ibm.ps2.hardware on Sun Aug 2 20:28:50 2026
    From Newsgroup: comp.sys.ibm.ps2.hardware

    The 9545 used the 9540 case and the 9533 sysboard.

    Josh Rodd wrote:
    Kevin Bowling <kevin.bowling@kev009.com> wrote:
    Could it be directly CPU bus attached? I would guess it would show up
    in a particular way if it was ISA attached.

    I have guessed that it was, but I haven't touched a PS/2 E in the flesh
    since the late 1990s.

    "Spiritually" I suspect it derives a lot from the Model 35 SLC design,
    which had a 84F7985 IBM SVGA which looks to be attached to a local bus.
    The PS/2 E was positioned to be "premium", hence why it had the faster
    XGA-2 versus the cheap SVGA chipset.

    (Didn't some Mod. 30s and Mod. 45s use the same planar as the PS/2 E?)

    With that said, this project probably isn't too relevant to the PS/2 E
    since it managed to be a lot faster than a typical 386SX and in
    particular has a much more capable video system. It's actually a bit of
    a genius design - the only thing slow are peripherals that are dog-slow
    anyway, like the IBM 54G0135 drive, spinning just over 3,600 RPM and
    running at about 1MB/sec.

    Stacked up next to an MCA bus ThinkPad 720, the PS/2 E's video system
    performed substantially better than the 720's MCA-attached WD video
    chip... the same cannot be said for the ESDI drive in the ThinkPad 720,
    probably because that drive could actually do DMA.

    Incidentally the PS/2 E was a tragedy when they were being liquidated at >>> Best Buy... 4MB machines with 1MB spent on the video subsystem, booting
    OS/2, and just swapping like crazy to a laptop-speed drive doing
    programmed I/O. A 71F7010 4MB upgrade made the thing actually usable.


    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Josh Rodd@joshua.at.rodd.dot.us@rodd.us to comp.sys.ibm.ps2.hardware on Sat Aug 8 03:18:18 2026
    From Newsgroup: comp.sys.ibm.ps2.hardware

    I seem to recall some Japanese thing (not sure if it was a PS/55 or if
    they were calling them PS/2s in Japan by then) that had both ISA and MCA
    slots, and it looked very much like a 9533 PS/2 E.

    Louis Ohland <ohland@charter.net> wrote:
    The 9545 used the 9540 case and the 9533 sysboard.

    Josh Rodd wrote:
    Kevin Bowling <kevin.bowling@kev009.com> wrote:
    Could it be directly CPU bus attached? I would guess it would show up
    in a particular way if it was ISA attached.

    I have guessed that it was, but I haven't touched a PS/2 E in the flesh since the late 1990s.

    "Spiritually" I suspect it derives a lot from the Model 35 SLC design, which had a 84F7985 IBM SVGA which looks to be attached to a local bus.
    The PS/2 E was positioned to be "premium", hence why it had the faster XGA-2 versus the cheap SVGA chipset.

    (Didn't some Mod. 30s and Mod. 45s use the same planar as the PS/2 E?)

    With that said, this project probably isn't too relevant to the PS/2 E >>> since it managed to be a lot faster than a typical 386SX and in
    particular has a much more capable video system. It's actually a bit of >>> a genius design - the only thing slow are peripherals that are dog-slow >>> anyway, like the IBM 54G0135 drive, spinning just over 3,600 RPM and
    running at about 1MB/sec.

    Stacked up next to an MCA bus ThinkPad 720, the PS/2 E's video system
    performed substantially better than the 720's MCA-attached WD video
    chip... the same cannot be said for the ESDI drive in the ThinkPad 720, >>> probably because that drive could actually do DMA.

    Incidentally the PS/2 E was a tragedy when they were being liquidated at >>> Best Buy... 4MB machines with 1MB spent on the video subsystem, booting >>> OS/2, and just swapping like crazy to a laptop-speed drive doing
    programmed I/O. A 71F7010 4MB upgrade made the thing actually usable.




    --- Synchronet 3.22a-Linux NewsLink 1.2