Technology shifts are defining business changes that ultimately will redefine the technology.
One thing that became clear at DAC this year is that the next big collision won’t be the technology itself. It will be the business infrastructure that supports the technology.
The integration of IP, software, subsystems, and ultimately entire die and packages will have a major impact on the chip industry on all levels. For some companies it will be good. For others it will be bad. But for everyone it will be disruptive.
There will still be a need for designs to take place at the system level, but what those designs entail could change significantly. Software will have to be developed that both connects to the metal, as well as a connectivity system. That connectivity layer may be an operating system, a real-time operating system, or some sort of middleware or virtualization layer.
Networking on the chip will become much more important, too, whether it’s through interconnected bus architectures or network-on-chip technology. So will routing of power, planning for physical effects and characterization of IP. And so will the development of components that can be standardized and re-used.
But what isn’t clearly defined yet is exactly how all of this stuff will be sold, how the pieces will be consolidated from a commercial standpoint, and who will actually wield control over what gets used, how it gets used and by whom. Will it be the foundries, the EDA companies, some distribution operation that doesn’t exist yet, or perhaps even the assemblers of chips? And who will be responsible when something goes wrong?
How these business components get defined will play a critical role in not just what’s disruptive, but how it’s disruptive and why. Ironically, it was the technology that changed the semiconductor business. Now it will be the business that changes the technology.
–Ed Sperling
 |
 |
 |
 |
 |
 |
 |
 |
 |  |
Leave a Reply