Aardvark DailyNew Zealand's longest-running online daily news and commentary publication, now in its 25th year. The opinion pieces presented here are not purported to be fact but reasonable effort is made to ensure accuracy.
Content copyright © 1995 - 2019 to Bruce Simpson (aka Aardvark), the logo was kindly created for Aardvark Daily by the folks at aardvark.co.uk
Please visit the sponsor!
Right now the world is suffering from an acute shortage of key semiconductor devices.
Bread and butter parts like the STM32 microcontroller are in very short supply, causing manufacturers to pause production of products and even bringing car assembly lines to a grinding halt.
In response to these shortages, billions of dollars are being poured into building new fabrication plants. In fact the degree of investment is quite staggering.
Unfortunately for those caught out by the shortages, these new plants are not expected to come online for at least several years. This means that lead-times for new orders are now being measured in years and prices for what little stock is available has gone through the roof.
However, what will the scene look like in five or six years' time?
Are we potentially setting the industry up for failure due to an over-supply of fabrication capacity and a resulting crash in component prices?
Obviously we need a lot of manufacturing capacity to catch up with the backorders currently being booked but what happens once that pent-up demand has been serviced?
Might it be that some of those plants end up sitting idle or even being mothballed?
Then there's the problem of advancing technology.
Many of the plants being planned and built right now are based on present-day fabrication standards and technologies. How relevant will they be to the industry in a few years' time? We've already seen a pretty rapid move from 14nm to technologies as small as 5nm in just a few short years. Will the plants being built today be outdated before they're even commissioned?
Fortunately the answer to that is probably "no".
Many of the semiconductor devices presently in short-supply are based on relatively old-tech and use fabrication techniques that have already been around for a decade or more. There's simply no need to use "cutting edge" technologies to make a $1 microcontroller chip that already consumes just a few milliwatts of power.
Those cutting edge super-small technologies are really the domain of far more complex devices such as the CPUs and GPUs found in modern computers and smartphones. The far simpler microcontrollers that the car and other industries need can be produced with 40nm fabrication technology.
Never the less, the effects of dramatically increased production capacity will undoubtedly push prices down for regular jelly-bean components as we reach the middle of the decade. That has to be good news for everyone, except those companies and investors who are presently pouring billions into building these new plants.
In the meantime, I'm actually having a bit of fun doing some recycling of old devices so as to extract their valuable components for current projects.
I ended up with a big box of flight-controller boards, each of which has an STM32 processor on it that would once have been worth less than a dollar but today could be sold for $20 or more.
How many of those companies currently processing e-waste realise that they probably have a small fortune in these parts sitting in their warehouses. The real gold and value in that e-waste is probably escaping them because they don't know what they're looking for.
Now there's an idea... I wonder if I should pay a visit to the local e-waste centre and see just how many valuable devices they're throwing away instead of recovering and reselling. That could be a great little earner for someone with an entrepreneurial attitude and some spare time.
Please visit the sponsor!
Have your say in the Aardvark Forums.