RISC OS news and information from Drobe


Beta! \| About us \| Contact \| RSS	Webspace \| Tech docs \| Downloads \| BBC Micro \| Gallery \| Wallpaper

Unique StrongARM NC in auction

By Chris Williams. Published: 27th Oct 2005, 23:33:18.

Like a RiscPC but steeped in Acorn history

A StrongARM powered prototype Acorn Network Computer has surfaced on an online auction house. The kit, known as a 'Fast NC' or 'Office NC', is being sold off by ex-Acorn engineer Piers Wombwell, who used the unreleased machine to port Java to RISC OS.

The NC, which has been gathering dust in Piers's attic, is essentially a 200MHz StrongARM RiscPC in a slim case. The computer has 16M of RAM, 2M VRAM, a 10baseT network card and RISC OS 3.7 in Flash ROM. It requires a network to boot, as it relies on a BootP server to give it an IP address.

Piers said: "I intended it to be my front room MP3 player, but I couldn't be bothered networking my front room. So I bought an Airport Express."

Network Computers that made it into production usually sport a slower ARM7500 processor. The set top box was launched in 1996, although the ground work for it began in mid-1994. RISC OS continued to be used past the Acorn break up in IPTV STBs.

Links
StrongARM NC auction on ebay, less than 2 days left and highest bid over 60 quid.

Discussion

Viewing threaded comments | View comments unthreaded, listed by date

Most interesting story about the Acorn history found in the link "mid-1994".

The story says about Acorn predicting the market towards the end of the nineties to be heading into network computers ruuning off a server tha houses the hard disk and loaded with software.

It appears from the story that Acorn were thinking a little too far ahead of times instead of the immediate market?

So what really happened, does anyone know?

Steve Harrington

Sawadee on 28/10/05 10:25AM [ Reply | Report ]

They got caught up in the spat between Oracle's Larry Ellision and Bill Gates. Ellision saw Acorn's set top box development and decided that a similar concept of a thin client Network Computer connected to large severs running his software would be the way to break the Windows hegemony on the desktop. He touted Acorn's technology as demonstrator just well enough to get Intel worried, and when they announced they would support the concept with an x86 based device, Ellison then promply dropped Acorn. Apparently leaving them well out of pocket for work on the next generation of NC's of which the StrongARM NC was one.

Acorn then took the existing ARM7 based units and tried to develop the concept as a cheap internet terminal, for use in kiosks and a consumer version sold by Bush. This actually worked quite well initially and I convinced a couple of friends to buy one. But the commitment from ISPs were lacking, and after being passed around several times, it was eventually dropped leaving the boxes unsupported.

Anyway thats my understanding, but I'm sure piers knows far more, if he's allowed to say.

druck on 28/10/05 11:25AM [ Reply | Report ]

In reply to Druck:

I had one of the Bush set top boxes about 5 years ago. Used it quite regulalry but decided I needed email in which I could use attachements. I finally stopped using it when I broke the remote control/flip top key pad.

Thinking about it I might have it still lying around the house somewhere.

Col1 on 28/10/05 11:32AM [ Reply | Report ]

You're right, Steve - that's a very interesting link. In fact the whole of that acorn history site looks to be a very valuable resource I wasn't aware of. I'm slightly surprised that it says the ARM Development System used an ARM2. I'd always thought that it was an ARM1 in there.

Looks like this NC might go for a decent amount. Over a day to go and it's already attracted some interest.

ninja on 28/10/05 12:56PM [ Reply | Report ]

Unless I am misreading the article, it says that ARM1's were used in the development machines

Walks on 28/10/05 1:34PM [ Reply | Report ]

I hope Chris Whytehead does manage to win it. He's currently the high bidder. As then it's likely more photos and info will appear on his Acorn Computer pages [link] He's always been helpful to me when I've had any questions.

flibble on 28/10/05 1:44PM [ Reply | Report ]

I hope Chris Whytehead does manage to win it. So do I : -))

But it would be a few months before it appears on [link] as I need to catalog 3 boxes of hardware upgrades etc., before putting them away, so I have room for more computers. Still got the A680 to add to the site.

chriswhy on 28/10/05 6:33PM [ Reply | Report ]

druck: The problem with supporting NCs on an ISP was the file system - it required NFS, which is not very well suited to dial up connections (slow, and it resulted in effectively keeping connections open, causing nfsd to suffer badly).

We supported them at ArgoNet - but that was only because I wrote a filing system that was much better over a dial up connection (well over twice as fast for the same data). We still had to use NFS to soft-load the new filing system, but otherwise it was fine (the NFS connection was terminated as soon as possible).

Other ISPs weren't interested in supporting them because they were a nightmare to set up. We were able to sell a few "ISP kits" around the world (ISTR Sweden, Finland and Philippines).

tribbles2 on 29/10/05 12:33AM [ Reply | Report ]

flibble: I've added some decent pictures of the thing so you don't have to wait for Chris: [link]

Secondly, I think druck's getting a little confused between DEC and NCI (Oracle). Acorn were contracted to NCI on a time and materials basis, so it's unlikely they would end up out of pocket.

Acorn were working with DEC on a StrongARM based STB chipset. SA1500 was DEC's StrongARM with an on-chip VLIW co-processor for MPEG decode. The SA1501 was Acorn's companion chip which did the video display - sort of like the IMOD but targetted at TVs.

After Intel bought DEC, they cancelled the SA1500 which meant the SA1501 was useless. That left Acorn considerably out of pocket.

I thought the ARM Development Systems were ARM1. My Springboard manual implies the Springboard shipped with ARM1s (no mention of MUL instruction in the assembler guide, and it has a code sequence for how to do a very slow multiply), and that's dated 17th July 1987. The BBC co-processor is considerably older.

Piers on 29/10/05 11:13AM [ Reply | Report ]

tribbles2: The next generation NC (which was initially to be marketed by RCA in the US) fixed the NFS problem with a new FS based on FTP (smurfs, it was called internally - I forget its official name)... Never finished or released, of course.

Piers on 29/10/05 12:00PM [ Reply | Report ]

The 'PC ARM Application card' (which I believe to be the Springboard hardware) is indeed an ARM1 (VL2333-QC). And the Springboard manual does list the entire ARM instruction set, which as you say, has no MUL).

wm on 29/10/05 2:53PM [ Reply | Report ]

Well done, Chris! Yet another computer added to the collection.

jymbob on 30/10/05 1:19AM [ Reply | Report ]

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

Create a new account
Forgot your password?

Search this website