You would have to be trying pretty hard to be unaware of the ongoing chip shortage that has left vehicles unassembled, computers unshipped, consumers concerned and OEMs panicking.
Triggered by COVID-19-related supply chain interruptions, chip manufacturers around the world are struggling to find the raw materials needed to build their products. While the industry is attempting to pivot, it’s nowhere near an instant process, and recommendations made even a few months ago can quickly become out of date as semiconductor manufacturers try to find their feet.
I’ve heard several different takes on when the chip shortage will end. The latest date I’ve heard was mid-2023, but Gartner VP for semiconductors and electronics, Gaurav Gupta, said that some sectors will see a return to normal by mid-2022. Gupta couches that prediction carefully, though, saying that mid-2022 is when we may see a balancing out of supply and demand but it won’t be a total return to normal.
“There are some devices right now, which have a very long lead time, 52 weeks or more. Then there are some devices which have a 28 week lead time, or 32. What we are anticipating by mid-2022 is for lead times to go back to the normal level,” Gupta said.
A six-month wait for a return to normal lead times (which equal out to additional waiting) may not be what many businesses, especially OEMs with dwindling supplies, want to hear. There may be ways to avoid having to suffer for the better part of another year, though.
How to buy hardware during the chip shortage
If you’re a business in the difficult place of nearing the end of a hardware replacement cycle during the chip shortage, sorry: There may not be a lot you can do to get new hardware until things level off. Rather than worrying about getting a new load of company laptops, Gupta said that it’s time to do what the rest of the world did during the pandemic and move to the cloud.
Gupta said that Gartner has seen a general trend of businesses that deploy laptops or desktops for employees adopting cloud services over the past year to reduce dependence on internal hardware. “This won’t be a temporary solution. We were seeing this trend coming and the pandemic accelerated that trend,” Gupta said.
If you’ve been holding off due to uncertainty about cloud services, Gupta said that the past year as proven it’s a mature technology ready for the big time. “People used to have concerns about latency, or data security. Now, with all the major hyperscalers that have been investing in cloud data centers and the like, I think those concerns have to a large degree been mitigated,” Gupta said.
There are businesses for whom adopting cloud technology in lieu of hardware upgrades won’t be realistic. For those businesses Gupta recommends ditching custom hardware orders and going for what’s in stock and ready-to-ship, which can turn a months-long wait into one that lasts a week.
How OEMs can fight the chip shortage
End-users and consumers have definitely been affected by the chip shortage, but the bigger impact has been on OEMs who rely on chips to build the products they sell. Combatting the chip shortage can take on a different form depending on what the company needs and the types of chips they ship.
Automotive companies, for example, have a problem because the types of chips they ship are existing, proven trusted components that aren’t cutting edge. “From a semiconductor perspective, the shortage is more on the legacy side in the mature devices,” Gupta said. That means that auto manufacturers, and other companies using older types of chips, are the ones actually being hit by silicon issues.
Other types of OEMs, who manufacture products with cutting-edge silicon like GPUs and CPUs, aren’t facing as severe a shortage because the demand for those newer types of chips is much smaller. Shortages in newer chips is actually being caused by shortages in the ABF substrate housings those chips sit in, Gupta said.
His recommendation to companies that rely on older style chips? Start ordering newer types of chips.
“You have to understand the source of the shortage today is with legacy and mature tech devices. It obviously makes a lot of sense to migrate your products to newer, more advanced devices,” Gupta said. Along with being easier to get ahold of, designing for newer silicon now means that your chip design will last longer into the future, saving costs down the road.
Smaller companies that lack lobbying power don’t have much leverage with semiconductor manufacturers, which Gupta said has made constraints even worse for them: While tech leaders are able to convince the government to ensure they receive an allocation of chips, other companies are left behind. “Work together, form consortiums and get support from the government to put pressure on chip foundries,” Gupta said, citing a similar successful move by the automotive industry earlier during the pandemic.
SEE: Checklist: Return to work (TechRepublic Premium)
Lastly, Gupta recommends that businesses bite the bullet and place a large, long-term order with a chip foundry. “What semiconductor foundries have started doing is prioritizing non-cancellable, long-term orders in order to weed out double and triple bookings,” Gupta said. This was a major cause of initial shortages and supply interruptions during the pandemic, Gupta said, because businesses started putting in panic orders that inflated demand. Flushing out false orders will be a major part of industry recovery for chip manufacturers, Gupta said.
“Chip manufacturing is a capital-intensive industry. They won’t just expand production capacity based on demand today, but if an OEM signs a long-term contract they’re more confident, and that means orders get made more quickly,” Gupta said.