| A9 public presentation tomorrow |
|
Published: 14th Dec 2004, 16:20:02GMT Source: drobe.co.uk
By Chris Williams |
| Page 1 of 1 |
|
| Some further details on A9home, plus screenshots |
|
AdvantageSix, aka STD, will be demonstrating an A9 machine tomorrow evening at the RISC OS North West User Group.
The machine boasts a mid-200MHz ARM 9 processor, Embedded RISC OS Adjust, 128M of RAM, CompactFlash storage, a graphics framebuffer capable of 640x512 and two PS/2 interface ports. The device will be packaged into a simple, vertical standing case, allowing users get a hands on experience with the machine. The drawfile rendering acceleration, powered by a recently developed AGP driver and closely linked to AdvantageSix, sadly won't be on show.
The kit for Wednesday night is based on models that they're shipping to larger, un-named OEMs to try out, and development work for these larger customers is, according to STD, benefiting what will eventually be the 'A9home' - a desktop version of the A9 series. STD remain tight lipped on exactly who their customers are, and as these are paying clients, their priorities are focused on them, rather than solely on the desktop market.
Embedded Adjust plays on the modularity of modern RISC OS 4, allowing STD to enable and disable OS modules and components as required - with the A9home having most or all of the components considered to be part of 'desktop RISC OS' enabled. Below are some screenshots of the desktop from the machine - and yes, they just look like the regular RISC OS 4 desktop, but it's reassuring in a way to see that the familiar desktop hasn't been entirely discarded in the embedded technology climate surrounding our platform.
 
 
There are currently no timescales or prices for the A9home yet, and STD are to later draw up what they think the A9home should feature and present the specification to the machine's retail dealer - currently rumoured to be CJE Micros. A desktop machine packing around 200Mhz may not inspire the majority of punters, but with faster ARM9 cores on test silicon rumoured to be in developers' hands, anything could happen. Possibly.
STD have denied that the A9's outing has any thing to do with Castle's recent MiMagic developments, adding that the user group simply asked them to demonstrate their new hardware in public. Particular developers have recently, quietly poured scorn on the MiMagic development work, claiming that it's not particularly newsworthy because Pace had the NeoMagic development board working with 32bit RISC OS in 2001.
Pictures and praise of the A9 have appeared in recent editions of Qercus and Archive.
Links
STD websiteThis article has been linked to, or is available in the following formats:
|
| |
|
Revin Kevin
14/12/04 4:45PM |
Nice to see new hardware being shown around in public. |
jc
14/12/04 5:34PM |
RONWUG has been asking Advantage 6 for a while for a date they would demo the A9 and members at the meeting two months ago were given the nod that there would be a demo in November or December. Stuart's holiday and Matt's cold last month made it a December demo!
Now that Drobe has given the meeting extra advertising I'm worried about the mince pies lasting! |
SimonC
14/12/04 6:08PM |
Wasn't sure if I was going to go (Wednesdays clash with other things), but if there are mince pies on offer I might change my mind... |
JWCR
14/12/04 10:11PM |
If memory serves me, the A9 is quite a compact board. Am I just hoping in vain that an A9home is just a few short steps away from an A9laptop. |
md0u80c9
14/12/04 11:59PM |
JCWR - The A9 appears to be a logical progression of a Unipod (or vice versa, the Unipod being a spin-off from the A9, I don't know). Given how many peripherals were mounted on the Unipod - both in terms of what was used and what wasn't, I would imagine that the board sizes are not dissimilar.
As to making it a laptop; I would imagine getting the components (cases being a well-known sticking point!) and sufficient market size to produce a laptop will probably be more prohibitive than the act of making one.
If an OEM were to order such a thing in sufficient quantities, then I suspect anything is possible; I would doubt that the RISC OS enthusiast market would draw sufficient numbers to be able to obtain the components :o(. Personally, given my current needs for a laptop, I'd have to agree I'd be on the front row for one  . |
martin
15/12/04 11:40AM |
There is a robustness and professionalism surrounding what Advantage Six do that inspires confidence.
I hope tomorrows presentation inspires someone sufficiently to write it up for drobe.
I'd like to join md0u80c9 on that front row for a laptop : just ramp the clock rate to the max, please. |
em2ac
15/12/04 12:41PM |
hmm glad to hear about the AGP development...now all we need is the PCI Express 
Why does the graphics buffer have such a low resolution? |
diomus
15/12/04 1:00PM |
emac2:
The low res is because they're using the SoC's on-board LCD controller, which is designed for small devices - as I understand it.
Chris. |
AMS
15/12/04 8:21PM |
In reply to Chris:
"Particular developers have recently, quietly poured scorn on the MiMagic development work, claiming that it's not particularly newsworthy because Pace had the NeoMagic development board working with 32bit RISC OS in 2001."
Nice, a good article about the A9 and it winds up slagging off Castle (do Castle make the A9, do they have anything to do with the A9 or did I miss something ?).
I'd point out Pace didn't release MiMagic - Castle did.
Look Chris if you want to "quote" people being negative how about "oh we had 200MHz ARM processors in 1995 and ones faster than that not much later and look 2004 and we're back to 200 MHz - wonderful....."
See if you look at things in the right way nothing is newsworthy |
dgs
15/12/04 9:07PM |
In reply to AMS:
Why criticise Chris for reporting on what embittered developers say?
One very well-known Acorn dealer told me (on the phone) that my Iyonix would probably have to be sent back to Castle because it was illegal. Maybe the developer that Chris overheard was the same person? I'd think that quite likely.
(I'd hope none of the people who complained about Castle's concerns over RISC OS licensing had any connection with the dealer in question!)
In the meantime, my Iyonix has gone back to Castle, but only for a disk upgrade, adding a second graphics card for dual display with Geminus, and adding a PCI-TV card. At the same time, I asked Castle to apply all the latest upgrades, and also bought a second Iyonix.
Personally, I think developers should spend less of their time being bitter, and more of their time getting on and releasing products that we want. The Iyonix has certainly achieved that, despite the nay-sayers!
Drobe never shys away from raising controversy, so why should we criticise Drobe just for pointing out that some developers have (sadly) been on a bad-mouthing spree yet again?
As for 200MHz, please remember the RiscPC was designed to support processors with maximum clockspeeds around 80MHz to 100MHz. The performance advantages of the StrongARM surprised many people, but there's certainly no reason to think the A9home shouldn't achieve performance far surpassing everything
except the Iyonix.
In short, you can claim things aren't newsworthy, but proving it is another thing.
dgs |
md0u80c9
15/12/04 9:43PM |
Or including Iyonix of course for that matter. Depends upon how fast the rest of the system responds, which processors are eventually used in the A9home etc etc.
We'll have to wait and see what the benchmarks pull up in the forthcoming year. |
mrchocky
15/12/04 10:02PM |
No, it's unlikely that any system based around any existing ARM9 will be as fast as an Iyonix. But you can be sure it will be rather faster than a RiscPC, even at 200Mhz. |
dgs
15/12/04 10:03PM |
In reply to md0u80c9:
Absolutely - if the A9home uses an ARM11 CPU, it will almost certainly outperform the Iyonix by a large margin!
But at the moment, it seems likely to use an ARM9 CPU.
As martin has already said, please add me to the waiting list for an ARM9 native RISC OS laptop without checking first. My RiscStation laptop deposit cheque is ready to be re-issued, with a one year deadline this time.
Oh, and I agree with martin about ramping up the clock speed as far as possible, too 
dgs |
AMS
15/12/04 10:06PM |
In reply to dgs:
Chris does a great job (and I publically thank him here and now for it), and far be it from me to critise him for what was essentially a generally a very good article. My comments may have read more "negative" than they were intended (hey look there's even a smiley at the end).
Unlike you (or me, or Chris for that matter) our comments can be publically seen and commented on (even negatively at times). It's easy for someone in the shadows to pour cold water on things like that "developer" has and never have to justify any of it - it is as, you said, their claiming something isn't newsworthy doesn't necessarily make it so.
As to the A9 (the subject of the article), yes it's only 200MHz but as I (and others) have said many times you can only compare clock speeds if the architecture and system around the CPU are identical. In this case they aren't so, yes I'd expect the A9 to perform better than a 200MHz SA based RPC (but that exact amount will be hard to gauge until the machine is released). If STD get the price right they could be on a winner, because at the moment the "budget" RISC OS sector is rather badly served.
Regards
Annraoi |
bucksboy
16/12/04 9:06AM |
In reply to AMS:
Surely the budget RISC OS sector is excellently served by the profusion of s/h RPCs now available, which also happen to be more expandable than any new machine short of an Iyonix (serial, parallel, USB, SCSI, ViewFinder, addn'l slices)?
As to clock speed, I agree that this is less important than other system features: VARPC emulators record relatively low MIPS figures when benchmarked, yet perform comparably to an Iyonix given a powerful system.
George |
em2ac
16/12/04 10:56AM |
In reply to diomus:
knew there had to be a reason |
JohnB
16/12/04 11:23AM |
I find the A9 Home to be a very exciting development and am keen to see exactly what 2005 brings us. At the moment we don't have a capable mid/low range machine available to purchase. The Iyonix is priced for what it is - the fastest machine we have currently. So, unless an ARM 7500 meets your needs, your only option is to purchase a second-hand tricked out 10 year old machine.
Also the speed of the ARM 9 that will power the A9 Home has not been decided. As was pointed out in "Lifting the lid on the A9" and again in this thread - ARM9's can range as high as 533Mhz. While I don't imagine that we'd be lucky enough to see a 533Mhz A9 Home, I'd be even more surprised if the final spec of the production model was 200Mhz. If for no other reason than to avoid problems similar to those that RISCstation faced - marketing a new machine with a lesser/same clockspeed as machines currently in circulation.
And to finish I'd just like to say I'm looking forward to seeing the mystery of the drawfile acceleration revealed - go STD/A6 ! ^_^ |
Revin Kevin
16/12/04 11:23AM |
So how did it go? |
Col1
16/12/04 1:49PM |
In reply to Revin Kevin:
It went very well.
|
| |
|
|
|
|
Bookmarks!
We have a directory of Companies, User groups and more in our bookmarks section, check it out!
|
|
