Drobe :: The archives
About Drobe | Contact | RSS | Twitter | Tech docs | Downloads | BBC Micro

PCI is go for RiscPCs, A7000s

By Chris Williams. Published: 15th Sep 2003, 15:53:09 | Permalink | Printable

Windfall stuffs even more into faithful old kit

Windfall.nl logoPCI is now possible on the RiscPC architecture, thanks to Windfall Engineering's latest product - PCIFace. Cramming an AGP interface onto the Acorn podule bus was clearly not enough, now RiscPC and A7000 users can fit a podule-bus-to-PCI bridge interface.

PCIFace appears to work in a similar way to the Viewfinder graphics card, namely by acting as a hardware layer in between an off the shelf, industry standard PCI card and the internal podule bus. Software developers can then write the driver software so that the fitted PCI card can be used from RISC OS. Each card is supplied with documentation and supporting code.

Windfall Engineer's John Kortink told us that the PCIFace card is suitable for individual users who want to produce a one off driver for a particular PCI card they might have their heart set on. The card is also just as suitable for 3rd party developers who can develop drivers for PCI cards and, for instance, resell PCIFace cards with the driver software to end users.

"PCIFace will be available in bulk (at bulk prices) for other developers to bundle with their particular PCI solutions, or they can simply refer end users to buy their PCIFace direct from Windfall and just ship their PCI card + driver", John explained to drobe.co.uk.

It's also worth noting that the programming API for PCIFace isn't compatible with the published Castle PCI API. This, of course, further widens the gap between the Iyonix and RiscPC architectures, unless someone develops a suitable software layer to convert one API into another.

"It's unlikely compatibility will be strived for now or in the future", John added when we asked about the possibility of PCI API compatibility.

"One reason is that PCIFace doesn't really need any OS support (memory allocation etc.) since every PCIFace will simply be (exclusively) allocated the 16 MB EASI space of the podule slot it's in (like any regular podule would)."

Two PCIFace cards are available, one with DMA capability and one without. The DMA enabled card can shift data at a top speed of about 8 MB/sec, according to Windfall.

The development of the PCIFace card buys the Acorn era kit some more time by enabling legacy RiscPCs and A7000s to embrace the PCI standard which only really recently arrived at the doorstep of RISC OS. We'd love to see something like Simon Wilson's gradually-getting-there Iyonix PCI TV card driver, but for the RiscPC, however there's the slight problem of there being no where to sensibly overlay the captured frames.

Even so, there's still a great many possibilities in terms of PCI cards that could be supported. A RiscPC with USB, AGP and PCI support? What more could possibly be crammed in? What PCI cards would you like to see supported on RISC OS?


PCIFace website PCI and RISC OS in detail

Previous: MSN clients parry October lockout
Next: RISC OS mailing lists paralyzed


Viewing threaded comments | View comments unthreaded, listed by date | Skip to the end

This is great news! I was just thinking about the possiblity of something like this after someone mentioned it in the comments of a Drobe article a couple of days ago. Well done, John.

It would be possible to do frame capture with a PCI TV card in a RiscPC, but live video would be a little more difficult. A modest-sized window of 320x240 requires a transfer rate of 7.5MB/sec, and that's assuming that it can copy directly into video memory. I wonder if this is a possibility. I would be happy to help convert my driver for use in the RiscPC.

Updated PCITV website at [link] -- Simon Wilson, Boulder, Colorado

 is a RISC OS Userksattic on 15/9/03 4:13PM
[ Reply | Permalink | Report ]

this has made me happy, someone thinking about a project before I have!


well done! but why not use the Castles API?

As far as the PCI TV Cards go, you can easily pick up an MPEGing card that MPEGS before the Computer recieves the Data! (so better resolutions are achived for recording)

 is a RISC OS Userem2ac on 15/9/03 6:02PM
[ Reply | Permalink | Report ]

Looks like JohnK is making the RiscPC into an Omega, can we have a DDR-RAM adaptor next please? ;)

-- C'mon, mod me down, PUNK!

 is a RISC OS Usersimo on 15/9/03 7:27PM
[ Reply | Permalink | Report ]

When do we get the overclocked Iyonix-like podule bus speeds, John? :)

 is a RISC OS Userimj on 15/9/03 7:31PM
[ Reply | Permalink | Report ]

em2ac: MPEGging the data on the card would help the recording speed but it would hinder live video as the computer must decode the MPEG stream to display it. -- Simon Wilson, Boulder, Colorado

 is a RISC OS Userksattic on 15/9/03 7:41PM
[ Reply | Permalink | Report ]

simo: We'll all need 8 slot backplanes with max size cases ;)

 is a RISC OS UserAndrewDuffell on 15/9/03 8:29PM
[ Reply | Permalink | Report ]

Here's a truly freakey thought, Iyonix has a podule bus (as well as it's PCI one), so would it be possible to use JK's PCIFace to allow another (extra) PCI card to an Iyonix.

Granted the PCI will only run at the Iyonix's enhanced podule speed, but still it might prove interesting (you could use it for slow PCI devices of a sort that wouldn't demand the speed of a full speed PCI slot).

Yes I know that it doesn't support the Castle PCI Api, but it works on an RPC (which doesn't have PCI at all) so perhaps the module(s) PCIFace use are 32bit neutral/hardware neutral in which case JK may have a broader market than he thought (Iyonix users needing an extra free PCI slot !).

What do you think ?


-- Annraoi

 is a RISC OS UserAMS on 15/9/03 8:42PM
[ Reply | Permalink | Report ]

In reply to kstatic Excellent news on the Iyonix pci tv card. Nice update to the website too. Thanks for your hard work Simon.

 is a RISC OS UserBoroboy on 16/9/03 12:18PM
[ Reply | Permalink | Report ]

PCI is go? The heck it is. John's being rather abrupt and giving unhelpful or plain useless answers to questions about PCIFace - we've absolutely no idea how to drive it, so whilst we would *LOVE* to start a PCI project, we can't even scope it or check the feasability as JK bizarrely appears not to want to sell this new thing he's created. I suggest that everyone pester him to bits if you actually want to see those of us who want to write drivers, have enough info to write drivers! What a silly situation.

 is a RISC OS Userimj on 18/9/03 7:25PM
[ Reply | Permalink | Report ]

Please login before posting a comment. Use the form on the right to do so or create a free account.

Search the archives

Today's featured article

  • Java and RISC OS
    Nick Brown explores the state of play and future options
     47 comments, latest by em2ac on 28/09/07 12:34AM. Published: 19 Sep 2007

  • Random article

  • Is it still cool to hate Intel and IE?
    Times change, enemies don't
     31 comments, latest by anon/ on 21/10/04 3:29PM. Published: 30 Aug 2004

  • Useful links

    News and media:

    Top developers:
    RISCOS LtdRISC OS OpenMW SoftwareR-CompAdvantage SixVirtualAcorn

    CJE MicrosAPDLCastlea4X-AmpleLiquid SiliconWebmonster


    RISCOS.org.ukRISCOS.orgRISCOS.infoFilebaseChris Why's Acorn/RISC OS collectionNetSurf

    Non-RISC OS:
    The RegisterThe InquirerApple InsiderBBC NewsSky NewsGoogle Newsxkcddiodesign

    © 1999-2009 The Drobe Team. Some rights reserved, click here for more information
    Powered by MiniDrobeCMS, based on J4U | Statistics
    "It gives a bad impression to anyone wondering who's going to be next to go when the web sites reporting the news have stopped reporting anything"
    Page generated in 0.1432 seconds.