Demand for chips is huge with chip prices jumping higher every month and shortages of key components such as memory halting product manufacturing lines. The Semiconductor Industry Association this week reported a record $35 billion in August sales -- a jump of 24% -- the thirteenth consecutive monthly increase over the year ago period.
John Neuffer, head of the Semiconductor Industry Association, said: "Sales in August increased across the board, with every major regional market and semiconductor product category posting gains... Memory products continue be a major driver of overall market growth, but sales were up even without memory in August."
A big jump in sales doesn't mean a big jump in units sold -- it means prices have risen sharply. It means higher prices for new servers, network equipment, consumer electronics, to put it simply: the entire digital fabric of our future -- including all of its promises of advanced AI, superior healthcare and a myriad other projects of technological progress -- will cost more; and there will be less of it. And there's no guarantee that prices will come down this time -- as they usually do.
This problem of chip shortages and higher prices is normally solved quite quickly by the chip industry.
Every boom spurs an over-investment of capital in new chip fabs and the resulting glut crashes prices and the bust cycle begins. But the availability of cheap chips creates new applications and new markets and new investments and new progress is made. This has been the economic cycle that underpins the innovative leaps and bounds of the technological miracles that enable our modern world.
The semiconductor industry does not get the recognition it deserves for its role as the foundational technology that has fueled the engines of innovation in every industry and in every market. Every two years the chips get twice as fast at half the cost. Sloppy software runs like a gazelle -- and faster chips makes for a a faster route to innovation of all types.
For more than 50 years the chip industry has been vital to our fast pace of innovation and in making technologies affordable on massive scales. But this time those cheap chips might be a memory and prices could remain high and shortages lengthy and even ruinous to some companies.
The chip industry is struggling with sub10 nanometer manufacturing and the challenges are becoming ever more expensive to overcome. The slowing of Moore's Law means stalled innovation: AI needs brute force computing power, so does scientific research such as drug discovery; and so do a trillion business processes. Computing costs are not zero even though programmers write code as if they were.
We don't have anything that can drive innovation at the same pace, and as broadly, as the chip industry. It means that the future will be delayed and the fruits of innovation will be far less affordable and shareable.