As technology has advanced, semiconductors have found their way into everything that requires computing power, from coffee machines to cars. But the manufacturing output for semiconductors has not kept up with this change.
The semiconductor industry has also been hit with an industry rotation in demand that it was never prepared to deal with.
This happened at the start of the coronavirus pandemic when automotive manufacturers scaled back semiconductor orders. Lockdowns meant they weren’t making enough cars, so they scaled down and battened the hatches.
Meanwhile, the demand for data centre, computing and home device semiconductors soared. Rather than finding themselves down on orders, semiconductor makers were all of a sudden making more semiconductors than ever before.
And then the automotive sector came roaring back.
Now, the semiconductor industry is in a state of disarray. Manufacturers are struggling to make enough chips in a situation we’ve called Chipageddon. This is compounded by the fact that silicon prices are soaring, making chips more expensive.
How long will the chip shortage last? The latest opinions don’t deliver good news – IBM says the chip shortage could last 2 years.
The president of IBM, Jim Whitehurst, has said that the current chip shortage could last another two years. Here’s what he said in an interview with the BBC:
“There’s just a big lag between from when a technology is developed and when [a fabrication plant] goes into construction and when chips come out. So frankly, we are looking at couple of years… before we get enough incremental capacity online to alleviate all aspects of the chip shortage.”
What Whitehurst means is it takes a long time to set up a chip fab before it can start producing chips. It takes 12-24 months typically, so you have a situation where even if a lot of fabs are being built, they won’t contribute for years.
The chip shortage is so severe that it has led IBM to look towards other ways to meet demand. “We’re going to have to look at reusing, extending the life of certain types of computing technologies,” says Whitehurst, “as well as accelerating investment in these [fabricating plants], to be able to as quickly as possible get more capacity online.”
IBM isn’t alone
There is a serious imbalance in the semiconductor industry, and this is a problem many companies are having to contend with.
For example, Ford cancelled shifts at two car plants earlier this year and said profits could be hit by up to $2.5bn due to chip shortages. Meanwhile, Apple announced it would take a $3 billion to $4 billion hit due to the global chip shortage.
However, the most telling story of the semiconductor shortage comes from Samsung.
Samsung is the world’s largest manufacturer of DRAM and the world’s fourth largest semiconductor manufacturer, and even they are experiencing shortages, having to delay the launch of the next-gen Galaxy Note until as late as 2022.
The fact that Samsung is experiencing a chip shortage when it manufactures its own chips tells us everything we need to know – the chip shortage is severe. It isn’t a small shortage at all – it’s an enormous shortage affecting everyone across the supply chain.
Unfortunately, it looks like the global semiconductor shortage will be around for a few years yet, and things could get worse before they get better.
The semiconductor shortage is the result of a catalogue of problems going back several years. Here are some of the highlights:
Intel is the world’s leading supplier of CPUs for PCs and data centres and in 2018 they caused a chip shortage with the troubled development of 10nm chips. Intel’s mistakes have led to a shortage in CPUs for computers.
Declining DRAM prices
DRAM is a computer’s main memory. In 2019 and 2020, prices for DRAM declined, causing the biggest players – Micron, Samsung and SK Hynix – to curb their output. This led to supply constraints when the coronavirus pandemic hit.
The global demand for chips has hit an all-time high. Data centres, computers, cloud services, augmented reality, 5G, connected devices and connected vehicles are fuelling demand. This is great for chip sales, but the industry can’t keep up.
The U.S. created a semiconductor shortage of its own making when they levied sanctions against several Chinese companies, including SMIC and Huawei. This exasperated the chip shortage, placing strain on domestic manufacturers.
Coronavirus pandemic and cancelled orders
During the coronavirus pandemic, demand for semiconductors soared in some industries (e.g. electronics) and dropped in others (e.g. automotive). When demand came back for “down” industries, demand didn’t drop for “up” industries, leading to a shortage.
We now have a situation where carmakers are battling the electronics industry for chips. There aren’t enough chips to go around and increasing manufacturing capacity is impossible without significant investment in new foundries.
The electronics super cycle is not going to end anytime soon because there are so many tailwinds, including self-driving cars, VR, AR, AI, 5G and space travel. So, we cannot expect demand to drop and the chip industry to catch up with itself.
To meet demand, we need new foundries. These take 12-24 months to set up. Many companies are already building new foundries, or they are boosting capacity at existing plants, which is good news for the long run.
In the here and now, manufacturers can meet demand for chips by partnering with an electronics component distributor like us. We specialise in the procurement and delivery of electronic components and parts (including semiconductors) for a wide variety of industries from the world’s leading manufacturers.
The semiconductor shortage has affected the entire manufacturing supply chain but our close links in the industry mean we have better access to chips than most. No promises, but we have an excellent track record across all sectors.
Get in touch with us for a chat about your needs. We’re here to help.
Call: 01904 415 415