Squeak On BugLab Platform?

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Squeak On BugLab Platform?

Laurence Rozier
... <a href="http://buglabs.net/products" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">Bugbase seems like a nice platform for Squeak - does anybody know what kind of performance to expect from  128M RAM on a ARM1136JF-S-based microprocessor?

TIA



Reply | Threaded
Open this post in threaded view
|

Re: Squeak On BugLab Platform?

timrowledge

On 5-Nov-07, at 12:26 PM, Laurence Rozier wrote:

> ... Bugbase seems like a nice platform for Squeak - does anybody  
> know what kind of performance to expect from  128M RAM on a  
> ARM1136JF-S-based microprocessor?

Well, depends a *lot* on the cpu speed (ought to be around 800MHz  
really, that's what I think the iPhone does) and I haven't been able  
to find anything that tells me.

The nice thing about the 1136JF-S is that it has a *hardware* vector  
float coprocessor, so one  of the annoying performance issues that has  
affected ARMs for years is gone. It doesn't have a very big cache  
though, but apparently it is intended that you can control some or all  
of the cache to hold important data or code instead of relying on a  
cache controller. That means you might be able to keep a lot of  
important data available at almost register access speeds; very useful.

It ought to be somewhat faster than my old 600MHz PXA80321 which  
managed Squeak at  respectable rate but really didn't cope with  
Morphic too well, partly because of the complicated way I had to  
mangle pixels on the way to the screen.

My guess is that it ought to do quite nicely for a tiny machine,  
especially with a direct framebuffer display setup.


tim
--
tim Rowledge; [hidden email]; http://www.rowledge.org/tim
Strange OpCodes: EF: Emulate Fireworks