This Rambus blog references an excellent article from Ed Sperling. “The basic business proposition for third-party IP, says Sperling, is that it’s cheaper, faster and less problematic to buy rather than build.”
Another well-done article from Sperling – followed by a series of articles about the advances in Quantum computing.
Sperling’s point regarding the resue of IP is particularly relevant in today’s environment as we are introducing a plethora of new computing architectures--while simultaneously accepting both the reality that it is becoming an almost impossible challenge of maintaining the historic trends of lithographic reductions in line widths, while also dealing with the apparently insurmountable challenges of commercializing 450 mm wafers.
So imagine the challenges to OEMs trying to squeeze out profits in this situation.
And it only gets worse as we go forward over the next 10 years in the transition from well understood Amdahl-type computing architectures based on binary data and into the eventual world of Quantum computers.
However, IMO the computer hardware OEMs will also be forced to limit the number of variables and risks they face as they move forward. And historically this path leads toward standardized memory products that sacrifice some higher levels of performance in return for predictable supply from multiple suppliers, universal interfaces, and acceptable levels of performance.
The more things change, the more they stay the same.