Technology is at a Grand Inflection Point. Maybe our civilization is. It sounds rather, well, grand, but we can’t overstate the importance of this juncture. In short, here’s the situation:
- Data is compounding at an exponential pace.
- There are, potentially, billions of dollars of value in insights from this ever-expanding universe of data.
- Computers – hardware and software – need to catch up to squeeze out more of these hidden insights, for profit or social good or both.
- Software is catching up to solve this problem. The real bottleneck is in hardware – processing power.
- During this time of exponentially growing data, Moore’s Law predicts a plateauing of processing power – based on the current laws of physics – because soon enough it become impossible to pack in more transistors in a square nanometer in any logic chip.
- The danger is that we’re approaching this “end” of Moore’s Law on the hardware side just when there’s a paradigm shift on the software side – AI or self-programming programs.
- Hardware needs to facilitate AI. Semiconductor companies are working hard to solve this problem in the face of Moore’s Law.
This is the dilemma: While investing in globally dominant logic semiconductor companies (and the ecosystem) seems like a slam dunk, there are 2 major challenges that keep us up at night: China, and Heterogenous Computing.
We want to be invested in semiconductors over the long term because we believe it’s one of the most important industries, with big implications in important fields such as autonomous vehicles and healthcare. But in terms of investments, are we’re playing with fire? In this analysis, we try to answer that question; we squeeze out insights – actionable insights – from both structured and unstructured data. But we do it the old-fashioned way – with natural intelligence (we hope). By the end of this analysis, we should have a clear prescription for our semiconductor portfolio – buy, hold, or sell.