Tech Trends – The Semiconductor Surge

The global semiconductor shortage has been affecting industries around the globe since its effects really started to be felt during the start of the Covid crisis. Now, it seems that we’ve finally reached the end of the four-plus year shortage, with major chipmakers such as AMD and Intel signalling that supply growth has finally reached a level to meet demand. In this issue of tech trends, we explore this latest development in depth, detailing the factors which caused the shortage in the first place, as well as major moves by industry players looking to take advantage of the end to the semiconductor shortage.

?

When analysing market movements as large as a general shortage, and coming out of said shortage, it’s a lot more complex than simply attributing both of these factors to one single element. It tends to be the case that various different market movements work in concert with each other to either create a shortage, or a surplus. However, one major factor which can be attributed to lifting global semiconductor supply out of its years’ long shortage would be the fact that this year, major markets for chips utilising semiconductors such as smartphones and PCs, have shrunk. This in turn is mainly due to a scaling back of purchasing power amongst consumers and organisations owing to a weak global economy coupled with rising inflation. What we are then left with, is a massive oversupply of chips, estimated to have caused a record loss for chip manufacturers to the tune of $12 billion. Thus, the global semiconductor and consequentially commodity chip shortage has swung around to the opposite end of the spectrum, where lower demand does not match higher supply.

?

A return to normality within this market is on the horizon, however, with demand in markets such as Smartphones and PCs slowly recovering (demand for PCs in particular jumped to only an 11% decline in the June quarter compared to 30% in the previous two quarters.) The AI market has also begun seeing an increase in demand for powerful chips, however it still only accounts for a comparatively small percentage of the overall demand within the global chip market. While small at the moment, AI is predicted to have a monumental effect on demand for the global chip market. This is because "Advanced AI servers have significantly higher leading-edge logic, memory and storage content versus traditional servers, and every incremental 1% penetration of AI servers and data centres is expected to drive $1 billion to $1.5 billion of additional (chip equipment) investment," Lam CEO Tim Archer said. On top of this, chip manufacturers are looking to scale up their high-end chip production, as these products are often used to support AI related chips themselves. Therefore, barring another pandemic such as Covid, we can expect the exponential growth of AI to seriously boost the demand and this production of high-end chips and the semiconductors that power them. Whether this itself will put more strain on the recovering global semiconductor supply still remains to be seen.

While this is good news for pretty much any industry that utilises semiconductors, it is important to recognise that a return to normal growth in the chip market is still progressing slowly. For example, Pat Gelsinger, Intel CEO, predicted that data centre chip sales will slowly decline into the third quarter of this year, before a recovery in the fourth quarter. On top of this, economic recovery in China has been notably slow. As China is the world’s largest chip buyer, this has an impact on the general global outlook. While this recent recovery news is definitely positive, technology leaders should still be aware that the general chip market is still very much in recovery, and plan accordingly.

?

For those bold enough, there is of course opportunity in the space where every other player seems to be exercising caution. AMD has taken this philosophy in stride, with their recent $400 million investment plan into India by 2028. Part of this plan involves the construction of the firm’s largest design centre in the Indian tech capital of Bengaluru. This centre is set to be a 500,000 square foot behemoth, taking the company’s office presence in the country to 10. This comes off the back of the Modi administration’s push to facilitate global investment into India’s own chip sector, with the aim to establish a global reputation for the country as a chipmaking hub. AMD’s design centre will apparently be open by the end of the year, with 3,000 new engineering roles to be generated over the 5-year investment initiative. This will all come as excellent news for Modi’s administration, with the country unveiling a $10 billion incentive programme back in 2021 for its chip sector. However, since announcement, this plan has not seen much traction owing to difficulties surrounding clearance for a firm to set up a manufacturing plant in the country. AMD is not the only chipmaker to begin setting up shop in India, with Micron also investing $825 million into a semiconductor testing and packaging unit based out of Gujarat. As the global chip market recovers, India may emerge as the next manufacturing centre for this technology, which will only stand to benefit the global supply of semiconductors owing to manufacturing diversity. It remains to be seen, however, if this speeds up the market’s recovery, and lowers semiconductor costs for the future. Technology leaders should be watching this space keenly, in order to take advantage of any surplus which is generated.

?

Finally, across the pond, U.S. lawmakers have been urging Biden to re-assess and potentially tighten the country’s export rules on AI chips to China, as the general industry has been lobbying Congress to leave the current rules as they stand. Republican Mike Gallagher, and Democrat Raja Krishnamoorthi, who both lead a committee focused on China, called for a strengthening of general export control rules which took effect last October. These rules essentially cut off China’s access to high-end AI chips manufactured by US firms. They do this by capping performance of said chips if they are exported to China, limiting their processing speeds as well as the speed at which they can communicate with each other. The lawmakers also suggested Congress looks at how Chinese firms can access high-powered chips in the cloud – as major firms such as Microsoft offer ‘chips for rent’ as part of their cloud offerings. Both Intel and Nvidia took this in stride, creating kneecapped chips for export. However, at least Nvidia’s chips still have the capability to be used in the creation of AI programmes, doing arguably little to slow the growth of China’s AI market. Nvidia has already pushed back against tightening the rules further, claiming restricting the sale of its chips to China further "would result in a permanent loss of opportunities for the U.S. industry." The general reaction from the chipmaking industry has been to increase lobbying activity towards the US government, including CEOs for Nvidia, Intel, and Qualcomm attending a meeting with administration officials in Washington to discuss the China policy. The impact on the global chip market any further restrictions on China may have, again, remains to be seen.

?

So, an arguably much better outlook for global semiconductor supply and in turn the global chip market. But how did the semiconductor shortage get so bad in the first place? As mentioned at the top of this article, it all started during Covid. Lockdown had billions of people around the world essentially stuck at home 24/7 for extended periods of time. This included work and study, which led to a boom in the demand for smartphones, tablets, and laptops (all of which use semiconductors). There was also a surge in online activity, from gaming to shopping. Platforms like Teams and Zoom exploded in growth, leading to larger data centres. Here, we have our first culprit: unprecedentedly high demand for semiconductors which began to put a strain on the global chip market.

?

On top of this, an increased demand for medical equipment utilising silicon chips coalesced with the roll out of 5g (a technology which has high semiconductor content) which increased the pressure. Finally, the US Department of Commerce’s Bureau of Industry published its Entity List – a record of people, governments, or other entities, which are subject to US license requirements on the export and trade of specified items. This, of course, includes US technologies such as semiconductors manufactured by US chip giants such as AMD. Of course, Chinese firms found their way onto this list, causing China to start stockpiling high-end chips. The real death-knell came when China’s Semiconductor Manufacturing International Corporation, a massive semiconductor fabricator, was placed on the entity list, causing many of the firms’ customers to shift orders to Taiwanese foundries which were already under pressure. The situation was not helped by local phenomena such as a massive winter storm in Texas, followed by an earthquake in Japan coupling with pandemic based logistics issues creating a delay in the shipping and processing of semiconductors around the world. All of this was compounded by a general shortage in the materials needed for manufacturing of these chips, as well as shortages in the supply of components and devices in general. In short, there was a coalition of various different factors which led to the general semiconductor shortage the world has experienced for the last 4 or so years.

?

The consequences of the last four years’ shortage have been felt globally, with consumers suffering under large price increases for tech across pretty much all end markets. This was not an insignificant increase either, with long wait times for products still being an issue despite consumers paying a premium. This was quite notable with the launch of the latest generation of Xbox and PlayStation consoles, with scalping due to supply shortage become almost a norm for a couple of years. Manufacturers were also negatively impacted, with the most notable effect of this being a shift in inventory practices. Vendor inventory strategies shifted across the board, as they began holding tighter inventories and shelving ‘just in time’ and ‘lean’ inventory practices during the shortage. Finally, it’s important to note the loss of market share which smaller companies suffered through owing to larger companies having the capital to still procure chips even at heightened prices. This will have had a detrimental effect on innovation within industry, as well as making a potential monopoly easier for certain firms.

?

The global semiconductor shortage kneecapped growth in a host of sectors, negatively impacted global industry, and stifled innovation. The worst part is it was arguably completely avoidable. There are important lessons to be learned going forward to avoid something like this happening again on a global level. A good way to prevent the collection of issues which led to this is to diversify manufacturing facilities. This seems to have already been taken up by firms such as Intel, who are looking to expand their fabrication footprint beyond geographical locations they have traditionally worked in. We’re also seeing a rise in industry collaboration, with Original Equipment Manufacturers like Apple working closely with companies within the Chip industry to streamline supply chains. This is being taken up by hyperscalers and automotive OEMs as well, signalling a cross-industry effort to avoid another shortage. OEMs are even going so far as to guarantee supply via long term agreements, such as facilitating pre-payments and co-investing into the chip industry. This is a step in the right direction, and we hope to see collaboration such as this scaling up for the greater good of the industry.