If those other competitors in the market were as "up" to the task as KEIL appears to be, why don't more people use their products now? Has anyone ever taken a census?

I'm of the opinion that ARM will use this as a means for shifting the emphasis in the MCU market from the fast, simple 8-bitter, to the less-fast less-simple, but cheap, highly integrated and therefore easy to use, ARM types. If you look at the offering made by Samsung and a number of others over the past three years in the ARM market, you find lots of different application-targeted versions, to which you directly connect FLASH, SRAM, and SDRAM, without any "glue" logic, and you're done, from the hardware standpoint. I remember one ARM7 version that cost $7 in modest quantity, supported a full complement of external FLASH, SDRAM, SRAM, and contained an ETHERNET MAC, 2 HDLC, a UART for a console, and Parallel I/O, among other things. It came in a low-cost PQFP-208. The ev-kit for it cost about $200 and included a full LINUX implementation.

They could take over the MCU market with that sort of an aresenal. I've seen other ARM's in SOHO switch/routers, and you'll find lots of 'em in cellphones. Almost every silicon vendor, aside from Arm's competitors, has one or two, if not many.

Going after the most widely-used MCU core on the market makes perfect sense. ARM doesn't sell the hardware, but they know that their customers, who do sell it, are pushing down the prices. They also know that their core is supported by the enormously flexible and totally free GNU tools. They also know that programmers like those tools. How could it hurt them to seize control of the most popular development tools for their main competitor?

RE