Intel have dabbled in many different segments and have always ended up dumping everything except their x86 line and associated motherboard lines.
Many of their explorations were bought in. eg. They bought the StrongARM stuff from DEC and then sold it to Marvel.
I would never design in an Intel FPGA or other part unless it has a second source.
Of course you should blame the company as a whole. It is the corporation that decides to buy/let ho whole divisions.
It seems many of these decisions to expand or contract are based on the current whims of Wall St.
"merging x86 cores and flash on a chip"... [I think you have that wrong]. They used to have XScale (which I believe is an ARM architecture) and divested themselves of it. I think your quote is wrong, and those "Hermon" and other Flash + logic chips had the ARM-based Xscale chip on them. Regardless, your point is better made by the fact that Intel COULD have had a large market share of ARM processors if they hadn't made that arrogant decision.
Intel is hobbled by the x86. Relative to register rich RISC-like architectures like ARM, register-poor CICS x86 instructions need a lot of clocking to get stuff done.
The only way to get x86 speed up is to vastly increase chip and instruction execution complexity. That means lots of transistors end up being clocked every instruction which means a lot of power gets consumed.
All up, that means x86 parts are expensive and power hungry relative to ARM parts on any apples-to-apples comparison.
High cost and high power is just totally incompatible with mobile space.
I agree. It seems to me that Intel has an "all-or-nothing" approach to markets.It has been selling in a huge market for years without any significant competition and this has resulted in complacency vis-a-vis the customers and competition. It keeps wanting customers who rely on it due to its pole and sole position, and small and insignificant competitors. Neither of this would be on offer in the mobile and future computing markets.
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...