But the turnaround is far from uniform. Everstream’s data shows that lead times for some advanced chips needed for medical devices, telecommunications, and cybersecurity systems, is around 52 weeks, compared to a prior average of 27 weeks.
Automotive companies that were badly affected by the pandemic because they initially canceled orders for components, were then blindsided by an uptick in demand, and had no spare inventory and little negotiating leverage when it came to ramping back up. Modern cars can have thousands of chips, and future models are likely to pack even more computing power thanks to more advanced in-car software and autonomous driving functionality.
“Anything automotive—or competing with capacity for automotive—is still highly constrained,” says Jeff Caldwell, director of global supply management at MasterWorks Electronics, a manufacturer of printed circuit boards, cables, and other electronics products. Actify CEO Dave Opsahl, whose company sells operation management software to automotive companies, says the supply of chips has not improved for carmakers, and in fact shortages of raw materials like resin and steel, as well as of labor, has also gotten worse.
Frank Cavallaro, chief executive officer at A2 Global, a company that finds, procures, and tests electronic components for manufacturers, says the current situation reflects the complexity of the chip market and supply chain. Many end products include numerous semiconductor components sourced from all over the world, and also require devices to be packaged by companies that are mostly in China. “It’s macro, it’s micro, it’s down to individual regions,” he says.
Gerdman of Everstream says the appearance of the new BA5 Covid variant in China has raised fears of draconian lockdowns that could hamper the production of chips and other products. She adds that uncertainty around future capacity as well as geopolitical restrictions on chip exports makes it difficult to plan ahead.
The geopolitical picture may well significantly increase global capacity to produce advanced chips. Legislation making its way through the US Senate would provide $52 billion in subsidies to increase domestic chip production. The US share of global chip production has fallen from 37 percent in the 1980s to 12 percent today. But while chip shortages have been cited by boosters of the subsidies, much of the money would go towards reshoring production of advanced chips. The country’s most advanced technology, from Intel, lags behind that of TSMC, presenting a potential weakness in America’s access to technology that promises to be vital for everything from AI to biotechnology to 5G.
The current downtown may only contribute to instability further down the semiconductor supply chain. “Unfortunately, a slowing economy brings with it the risk of some suppliers going into financial distress or liquidity crunch if they cannot access capital,” says Bindiya Vakil, CEO of Resilinc, a company that sells AI-based supply-chain management tools. “This can introduce a lot of risk into the supply situation. Companies should really monitor supplier financial health and collaborate closely with suppliers to give them favorable payment terms, upfront payments and so on, to help them with liquidity.”
The cyclical nature of the semiconductor industry even has some, including Syed Alam, who leads the global semiconductor practice at consulting firm Accenture, envisioning the shortage turning into a glut. “A rising concern for 2023 is the possibility of overcapacity for chip production,” he says. “Companies need to be focused on building an agile and resilient supply chain for the longer term, and be prepared to react.”
Like many nerds before me, I spent a goodly portion of my life searching for the perfect computing system. I wanted a single tool that would let me write prose or programs, that could search every email, tweet, or document in a few keystrokes, and that would work across all my devices. I yearned to summit the mythic Mt. Augment, to achieve the enlightenment of a properly orchestrated personal computer. Where the software industry offered notifications, little clicks and dings, messages jumping up and down on my screen like a dog begging for a treat, I wanted calm textuality. Seeking it, I tweaked. I configured.
The purpose of configuration is to make a thing work with some other thing—to make the to-do list work with the email client, say, or the calendar work with the other calendar. It’s an interdisciplinary study. Configuration can be as complex as programming or as simple as checking a box. Everyone talks about it, but it’s not taken that seriously, because there’s not much profit in it. And unfortunately, configuration is indistinguishable from procrastination. A little is fine but too much is embarrassing.
I spent almost three decades configuring my text editor, amassing 20 or so dotfiles that would make one acronym or nonsense word concordant with another. (For me: i3wm + emacs + org-mode + notmuch + tmux, bound together with ssh + git + Syncthing + Tailscale.) I’d start down a path, but then there’d be some blocker—some bug I didn’t understand, some page of errors I didn’t have time to deal with—and I’d give up.
A big problem I had was where to put my stuff. I tried different databases, folder structures, private websites, cloud drives, and desktop search tools. The key, finally, was to turn nearly everything in my life into emails. All my calendar entries, essay drafts, tweets—I wrote programs that turned them into gigs and gigs of emails. Emails are horrible, messy, swollen, decrepit forms of data, but they are understood by everything everywhere. You can lard them with attachments. You can tag them. You can add any amount of metadata to them and synchronize them with servers. They suck, but they work. No higher praise.
It took years to get all these emails into place, tag them, filter them just so. Little by little I could see more of the shape of my own data. And as I did this, software got better and computers got faster. Not only that, other people started sharing their config files on GitHub.
Then, one cold day—January 31, 2022—something bizarre happened. I was at home, writing a little glue function to make my emails searchable from anywhere inside my text editor. I evaluated that tiny program and ran it. It worked. Somewhere in my brain, I felt a distinct click. I was done. No longer configuring, but configured. The world had conspired to give me what I wanted. I stood up from the computer, suffused with a sort of European-classical-composer level of emotion, and went for a walk. Was this happiness? Freedom? Or would I find myself back tomorrow, with a whole new set of requirements?
When Russia annexed Crimea in 2014, the world’s chipmakers were even more dependent on Ukraine because the country supplied around 70 percent of neon gas. “There were delays in shipments because of border crossing issues,” says Shon-Roy, and the raw materials needed to make neon were also in short supply. “Russia was focusing a lot of their efforts on war and not making steel.”
Burned by that experience, the chip industry scrambled to diversify its supply. A company called Cymer, which is owned by Dutch chip giant ASML and makes the lasers used to draw patterns on advanced semiconductor chips, tried to reduce its consumption of neon. “Chipmakers are concerned about recent escalation of neon prices and supply continuity,” David Knowles, vice president and general manager of Cymer, said at the time, without specifically mentioning Ukraine.
Bondarenko says the price spike in 2014 was mainly caused by a feud between rival neon producers Cryoin and Iceblick, which is no longer operating. However, if access to Russian crude does become an issue, she says, Cryoin has enough supplies to keep production going until the end of March. If that runs out, she claims there are Ukrainian crude producers that Cryoin can turn to as alternatives.
Instead she is more worried about getting neon out of the country. “Borders right now are very overloaded as people, civilians, are trying to evacuate,” she says. “If the authorities of countries where our clients are located are able to influence the border situation for the commercial shipments then that would be a great help [and] it will not affect the whole industry worldwide.”
Chipmakers have played down how much they will be affected by the crisis in Ukraine. “There’s no need to worry,” Lee Seok-hee, CEO of South Korean chipmaker SK Hynix, said last week, adding the company had “secured a lot” of materials. Koichi Hagiuda, the minister of economy, trade, and industry in Japan, said Japanese chipmakers are not expecting a “major impact” on their operations because they can source materials elsewhere. The country imports 5 percent of gases used in semiconductor production from Ukraine.
But there are signs that despite the warning of 2014, Ukrainian neon still plays a major role in the industry. ASML told WIRED it sources “less than 20 percent” of the neon it uses in its factories from Russia or Ukraine. “Along with our supplier we are investigating alternative sources in the event of a supply disruption from Ukraine and Russia,” a spokesperson says.
There are concerns that the US is even more vulnerable. Last week, the White House urged US chipmakers to find alternative suppliers, Reuters reported. “We see huge amounts of imports coming into the US from [Russia and Ukraine],” says TechCet’s Shon-Roy. “It is my educated assessment that what’s coming into the US from Russia and Ukraine could be as much as 80 to 90 percent of all [neon] imports.” US chipmaker Intel did not respond to a request for comment.
But sourcing neon from elsewhere will not be easy. Any disruption in Ukraine will hit chipmakers at a time when the industry is already under intense pressure from post-pandemic demand. “The drive behind increased production is so strong that it is causing strain in the supply chain everywhere, even without a war,” Shon-Roy adds. “So there is no excess supply of this kind of gas that I know of, not in the Western world.”
More Great WIRED Stories
The Federal Trade Commission has sued to block Nvidia’s acquisition of Arm, the semiconductor design firm, saying that the blockbuster deal would unfairly stifle competition.
“The FTC is suing to block the largest semiconductor chip merger in history to prevent a chip conglomerate from stifling the innovation pipeline for next-generation technologies,” Holly Vedova, director of the FTC’s competition bureau, said in a statement. “Tomorrow’s technologies depend on preserving today’s competitive, cutting-edge chip markets. This proposed deal would distort Arm’s incentives in chip markets and allow the combined firm to unfairly undermine Nvidia’s rivals.”
Nvidia first announced its intention to acquire Arm in September 2020. At the time, the deal was worth $40 billion, but since then, Arm’s stock price has soared, and the cost of the cash and stock transaction has risen to $75 billion. The FTC lawsuit threatens to scuttle the deal entirely.
“As we move into this next step in the FTC process, we will continue to work to demonstrate that this transaction will benefit the industry and promote competition,” an Nvidia spokesperson told Ars. “Nvidia is committed to preserving Arm’s open licensing model and ensuring that its IP is available to all interested licensees, current and future.”
The FTC isn’t the first government regulator to scrutinize the deal. In October, the European Union announced that it was investigating the acquisition, and last month UK officials said they were concerned that the merger would threaten both competition and national security. China’s regulators are also looking into the deal, Nvidia said.
Much of the angst surrounding the acquisition stems from the fact that, for most of its history, Arm has been a relatively neutral player in the semiconductor world, offering access to its intellectual property to nearly anyone willing to pay the licensing fee. In the complaint, the FTC called Arm the “Switzerland” of the semiconductor industry. Customers fear that an Nvidia-controlled Arm would place them at the mercy of a competitor, while regulators are concerned that the acquisition would threaten to topple a massive, thriving ecosystem that depends on the architecture.
Arm began as a niche semiconductor designer, offering low-power chips for embedded systems and for portable devices like the Apple Newton and Palm Pilot. Over the years, as the performance of ARM chips improved and the importance of energy efficiency grew, the semiconductors found their way into a wider range of devices.
Today, Arm’s designs and instruction sets are widely used, appearing in everything from mobile phones to servers, automotive airbag controllers, and supercomputers. Recently, ARM chips have started making inroads in the PC world, ramping up pressure on incumbents Intel and AMD. Apple’s M1 chips showed just how competitive the architecture could be with x86 designs, and others have begun to follow suit. Earlier this week, Qualcomm announced a new Snapdragon processor, the awkwardly named 8cx Gen 3, which would run an ARM-specific version of Windows.
Because the ARM architecture is low-power and available to so many different companies, the chips have taken over the industry. Last year, companies sold 25 billion ARM chips, a more than fourfold increase since 2010.
Nvidia has also become an increasingly powerful player in the semiconductor world. Its graphics cards became key tools in machine learning and other artificial intelligence applications, and soon the company started selling chips tailored to high-performance computing. Its mobile Tegra chips, which license Arm’s designs, powered a series of smartphones and Tesla infotainment systems in the mid-2010s, and today they run inside Nintendo’s Switch.
Intel has spent the last few years lurching from one misstep to another, and even had to outsource the manufacturing of its latest chips to one of its biggest rivals.
Now, in order to recapture its former glory, the company is betting it can execute a series of tricky manufacturing shifts. But it’s also hoping that a rebranding campaign will convince people that it isn’t so far behind the competition after all.
Intel CEO Pat Gelsinger laid out a roadmap for several generations of chips at an event Monday. It includes new technologies designed to help the company compete with TSMC, a Taiwanese chipmaker that currently makes the most advanced and high performance computer chips, as well as Samsung in South Korea. The roadmap includes a timeline that will allow executives—and outsiders—to measure Intel’s progress.
In an early sign of success, Intel said Qualcomm and Amazon had agreed to be customers for its new foundry business, where Intel will manufacture chips for other companies; Intel said it will begin making chips for those companies in 2024. Gelsinger had announced plans for the foundry business in March, shortly after he rejoined the company where he was once CTO. However, in an embarrassing measure of how the company has fallen behind, Intel also plans to outsource the manufacturing of its most advanced chips to TSMC.
Gelsinger said Intel will adopt a new naming scheme for coming generations of chips. Currently, chipmakers refer to new chipmaking processes or “nodes” using a nanometer scale, with Intel currently using what’s known as a 10-nanometer process and TSMC using what it calls a 5-nanometer process.
The nanometer scale once referred to the actual size of a transistor gate, with continued shrinkage guaranteeing better performance. (A nanometer is one-billionth of a meter; a human hair is 50,000 to 100,000 nanometers thick.) One of Intel’s founders, Gordon Moore, famously stated in 1965 that progress in chipmaking could be measured by the ability to shrink roughly twice as many transistors onto a chip every two years.
But the nanometer scale no longer refers to actual distances on a chip, and Intel and others say that its current chips perform like those made on TSMC’s 7-nanometer process. It plans to adopt a naming scheme that reflects this, with a new version of its 10 nanometer due this year called “Intel 7” that the company says will deliver 10 to 15 percent better performance per watt of power. The generations beyond that, to come in 2023 and 2024, will be called “Intel 4” and “Intel 3.”
“There’s always the question of where the marketing ends and where the engineering begins, but this is grounded very deeply in engineering reality,” Gelsinger told WIRED ahead of Monday’s announcement.
Stacy Rasgon, an analyst with Bernstein Research, says the technical roadmap presented by Gelsinger seems promising but it will ramp up pressure on the company to execute. “This is all great but the danger is going to be they stick their neck out and it goes wrong again,” he says.
Intel made a series of blunders under its previous leadership. The company was slow to adapt to the shift to mobile computing, which saw it lose market share to Arm, which makes blueprints for energy efficient chips used by companies including Apple, which uses Arm-based chips for the iPhone, the iPad, and some Macs.
Intel was also caught off guard by the rise of artificial intelligence. Nvidia, a “fabless” chip company capitalized on this trend with chips specialized for AI computations. Nvidia overtook Intel by market capitalization in July 2020.
On the manufacturing side, Intel was slower than TSMC to adopt the latest method for etching features into silicon, known as extreme ultraviolet lithography (EUV). Monday, the company said it would ramp up use of EUV, and had secured the first next-generation EUV machine from ASML, a Dutch company that is the only maker of EUV machines. The initiative will be costly, because each EUV machine costs around $120 million.