FSMLabs is finally able to discuss the single core cell phone handset solution we developed with Infineon Technologies last year. This system is cool technology – literally – since it allows parts count reduction on the handset and that reduces power consumption and that reduces heat production. More to the point, in a volume driven business, removing some hardware components is big news. I hope that some of the “no software royalties” hard liners will now pause and consider why paying royalties (or a production license) in order to not have to ship chips and batteries for those chips may be, you know, profitable.
The technology is very interesting. Essentially, we squeeze a lot of performance out of an ARM9 – making it jump back and forth between user interface, radio control, and wireless stacks. Turns out, the ARM9 is up to the job – it was being held back by poorly designed software.
There have been many product announcements of the imminent release of Linux based single core cell phone solutions. I think that in this industry people sometimes forget that there are hard technical problems. Just having a visionary CEO say “make it so” and let loose the loudly singing marketing flock doesn’t actually get the bits to move right. Our software worked because of the smart engineering effort of Cort Dougan, Michael Barabanov, and Zwane Mwaikambo and the engineers they supervised and the work of the engineers at Infineon and also the excellent evaluation work done by Infineon engineering management that allowed them to avoid the mistake that everyone else made in searching for paths to this product. (of course there is a lot more to the phone than FSMLabs work, but RTLinux is agile enough to move the processor to the right place at the right time.)