Speaking out of turn, but... > some thoughts; > - the 20 MHz might be artificial because I saw presentations > from ARM talking about 100 MHz for the Cortex M3 20MHz is the first set of devices from Luminary. They will probably publish a roadmap. > - the core is VERY new, meaning this is the first > implementation in silicon available, expect some debugging of > the core and an Errata Sheet coming up soon!? There are errata sheets. The core is from ARM, so it's well tested one assumes. > - Once a year has gone, there are going to be more products > using this core with the first problems debugged, so for now, > watching it is just fine, using a proven ARM7 core is our > better current option. Perhaps, but the Cortex-M3 is designed for low power, the core taking 1/3 the power of a comparable ARM7. And it's cheap and will probably get cheaper. > - May be the biggest issue, it is not "ARM" compatible, only "Thumb" > compatible. You need to touch any software you wrote for an > ARM micro because this core does not support ARM mode and by > definition of all previous devices any entry in exceptions > (interrupts) are in ARM mode. Although not binary compatible, Thumb-2 is pretty much an ARM instruction set... It only affects you *if* you use assembly language. > Cortex is probably the core to go with in the future, the > question remains when to start using it. Want to be a Guinea pig ;-) One assumes that there will be other vendors offering Cortex. Note that the Stellaris has the single wire debug and single wire trace facilities which ARM cores do not have. The Stellaris family is nice, Cortex-M3 is a nice little core. -- Paul.
Message
RE: [lpc2000] Re: Something different...
2006-03-27 by Paul Curtis
Attachments
- No local attachments were found for this message.