Iyonix 26 bit emulator in development

By Chris Williams. Published: 20th Oct 2002, 20:36:03 | Permalink | Printable

"omg, omg 26bit emulator on a 32bit system"

drobe.co.uk has today learnt that a 26 bit emulator for Castle's recently unveiled 32bit XScale computer is in the works.

Aemulor, pronounced aim-you-lore if you're one of those odd people who gets all upset over odd product names, presumably enables Iyonix users to run 26bit only software on the 32bit system. The full 32bit issue is discussed here, but essentially current 26 bit software will not work correctly on Castle's new 32bit machine unless it is made 32bit compatible or an emulator is employed to do the conversion on the fly. The Aemulor website is currently owned and maintained by spellings.net, which was recently mentioned in the Mail on Sunday newspaper.

Just before you question our editorial decision to publish this Aemulator 'news', we do know it's a typical "Coming soon" thing and we're painfully aware that the recent Castle announcement has actually raised more questions than answered. However, it is clear that some serious developments are afoot and we'll bring you news as it goes public. Stay tuned.

Links

Aemulor website
ARM on ARM emulation What's 'omg' mean?

Previous: chocky.org at South East Show
Next: Playing a winning hand with Vingt et Un

Discussion

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

Um, maybe they'll sell Aemulors to Microdigital. -- Spriteman.

Spriteman on 21/10/02 9:33AM
[ Reply | Permalink | Report ]

I've heard that a number of the 400MHz Xscale instructions are pretty slow compared with a 200MHz StrongARM.

Hence all those PocketPC owners complaining of a slower OS on their device even though they've gone out and bought a 400MHz Xscale device rather than a 200MHz StrongARM.

So, will a 26bit emulator be really slow?

quatermass on 21/10/02 10:05AM
[ Reply | Permalink | Report ]

I understand that while XScale is generally a little slower than StrongARM (at same clock speed), the main problem on the WinCE devices is lack of optimization for the XScale, and if this were done, XScale devices would be a fair bit faster than SA ones.

thegman on 21/10/02 10:17AM
[ Reply | Permalink | Report ]

Won't the DDR RAM make a huge difference anyway though? -- Jess

jess on 21/10/02 2:09PM
[ Reply | Permalink | Report ]

DDR helps to limit the problem.

xScale uses a longer pipeline compared to SA-110 (8 rather than 5 stages). Thus if you mispredict a branch (and flush/refill the pipeline) the penalty for such a miss on xScale is larger (I'd guestimate 8 to 5 worse !).

Careful use of conditional codes or branches can limit the "hit" you take. On the other hand if you don't do this then the performance you acchieve will not quite be as high as the difference in clock rate might suggest.

To further complexify things xScale has a larger cache, and a small "data scratch pad" area that may allow a speed improvement not available to SA-110.

The tall and short of it is that xScale only really gives you the improvement you'd expect if code is optimised for it, otherwise the performance will be a lot closer to StrongARM (or put another way you'll need an xScale with a clock rate substantially more than a 300MHz SA to compensate if you don't optimise).

Regards -- Annraoi McShane,

AMS on 21/10/02 5:23PM
[ Reply | Permalink | Report ]

Why is there any need to sell Aemulor to MD? They already have hardware capable of running 26-bit code natively.

Aemulor will be highly optimised for XScale, although admittedly the XScale's long pipeline isnt ideal for the kind of things an emulator does, but we will do the best with the hardware we've got!

Without giving too much away, remember that we are not just talking about an 233MHz XScale running in a RiscPC - this is an entirely new machine!

Regards,

Neil www.aemulor.com

spellinn on 22/10/02 2:28PM
[ Reply | Permalink | Report ]

Hi Neil,

will Aemulator be "bundled" with the RO5 (and a license paid by Castle), or will it be up to "end users" to source and pay for the system ?

I don't mind paying for an emulator, but I think perhaps from a sales/marketing viewpoint that perhaps having the thing bundled is the way to go (even if it means the machine is a few quid dearer to pay for it).

-- Annraoi McShane,

AMS on 22/10/02 4:03PM
[ Reply | Permalink | Report ]

Castle arent planning to "bundle" the full version, although they may include a demo copy, to allow you to test whether your legacy application will actually run under emulation. (We will also have a web database for this purpose also, so you can see before purchasing an Iyonix the status of your applications)

From Castle's perspective (and ours actually), they would rather encourage people to upgrade to the 32-bit versions where available - why should someone who only uses software which is available in 32-bit pay for an emulator bundled with the machine when they dont need it?

Those people that have apps which either take along time, or will never be updated are the ones who will need Aemulor, and hence should be the only ones paying for it.

Yes, from our sales persepective it would be great if Castle licensed a full version on every machine, however I think this is counterproductive to encouraging 32-bit conversion and development and moving the platform forward.

Regards,

Neil www.aemulor.com

spellinn on 22/10/02 9:39PM
[ Reply | Permalink | Report ]

I agree with Neil, though I think a lot of people will buy Aemulor till the 32bit developments catch up, or they will run their old machines side-by-side.

-- Simon Wilson, Boulder, Colorado

ksattic on 23/10/02 7:43AM
[ 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

South East 2008 show photos
If a picture tells a thousand words, here's a short story

8 comments, latest by hzn on 22/10/08 10:03AM. Published: 20 Oct 2008

Random article

RISC OS for Linux: New features from taskmanager to anti-aliasing

Discuss this. Published: 6 Jan 2001

Useful links

News and media:
Iconbar • MyRISCOS • ArcSite • RISCOScode • ANS • C.S.A.Announce • Archive • Qercus • RiscWorld • Drag'n'Drop • GAG-News

Top developers:
RISCOS Ltd • RISC OS Open • MW Software • R-Comp • Advantage Six • VirtualAcorn

Dealers:
CJE Micros • APDL • Castle • a4 • X-Ample • Liquid Silicon • Webmonster

Usergroups:
WROCC • RONE • NKACC • IRUG • SASAUG • ROUGOL • RONWUG • MUG • WAUG • GAG • RISCOS.be

Useful:
RISCOS.org.uk • RISCOS.org • RISCOS.info • Filebase • Chris Why's Acorn/RISC OS collection • NetSurf

Non-RISC OS:
The Register • The Inquirer • Apple Insider • BBC News • Sky News • Google News • xkcd • diodesign

© 1999-2009 The Drobe Team. Some rights reserved, click here for more information
Powered by MiniDrobeCMS, based on J4U | Statistics
"...a somewhat shallow article intended only to generate some journalistic controversy"
Page generated in 0.0346 seconds.