1 year on and now we find there is a very different scenario

the following from the FT in direct breach of their copyright law:

US chipmakers reel from sharp boom to bust Growing number of companies have issued weak sales and profit projections US chipmakers reel from sharp boom to bust on twitter (opens in a new window) US chipmakers reel from sharp boom to bust on facebook (opens in a new window) US chipmakers reel from sharp boom to bust on linkedin (opens in a new window) US chipmakers reel from sharp boom to bust on whatsapp (opens in a new window) Share Save Richard Waters in San Francisco 6 HOURS AGO 32 © FT montage/Dreamstime/Unsplash If Wall Street had any doubts about the speed with which the chip industry’s boom has turned to bust, unexpectedly gloomy financial forecasts from companies like mobile chipmaker Qualcomm should have put them to rest. “It’s kind of an unprecedented change over a short period of time,” Akash Palkhiwala, the company’s chief financial officer told analysts this month. “We went from a period of supply shortages to demand declines.” Qualcomm has sliced 25 per cent from its revenue guidance for the current quarter as weaker consumer spending hit smartphone sales. The forecast came as some of the leading chipmakers issued surprisingly weak sales and profit projections and signalled a round of job cuts ahead. Among those to take an axe to their forecasts, AMD warned that sales of processors for PCs this quarter would be down 40 per cent from last year, with profit margins also surprisingly weak. Intel, which cut revenue forecasts again after a big reduction the previous quarter, signalled thousands of job lay-offs ahead with a plan to cut as much as $10bn from its costs by 2025. A year ago, as share prices peaked, it was easy to believe that the chip industry had entered a new era. Giant new markets were opening up “from mobile to cloud to electric vehicles to metaverse,” Sanjay Mehrotra, chief executive officer of memory chipmaker Micron, said at the time. Supply chain problems had caused widespread chip shortages, buoying prices. Asked in an interview with the FT whether the chip business was still vulnerable to the sort of vicious cycles that have hit it in the past, Mehrotra declared: “Our industry is different, and certainly Micron is very different.” But less than a year later, the chip cycle has returned with a vengeance. In September, Micron warned that its revenue this quarter would tumble to $4.25bn, down 45 per cent from a year ago. The company also said its gross profit margin would collapse from 46 per cent to 25 per cent. In response, it slashed nearly half from its planned capital spending next year. By early October, the Philadelphia semiconductor index had fallen 47 per cent from its peak, compared with a 26 per cent fall in the wider market. But since then, as chip companies have confirmed investors’ worries about the depth of the downturn, the index has rebounded 16 per cent on hopes that a bottom to the sharp cycle may be coming into sight. Recommended ‘Tech cold war’: rising tensions in the critical chip industry How the US chip export controls have turned the screws on China For Wall Street, the severity of the sudden cyclical downturn has even overshadowed last month’s action by the US to block sales of advanced chips and chipmaking equipment to China. That move, which is likely to dent the industry’s long-term sales growth, was barely mentioned on the latest round of earnings calls. The intensity of the recent downturn owes much to an inventory glut that built up with surprising speed. Booming demand for many digital products and services during the pandemic fed the optimism of chip executives like Micron’s Mehrotra, leading to predictions of a strong period of secular growth ahead. At the same, chip shortages led to a deliberate build-up in inventory levels to guard against future supply shocks. That left the industry vulnerable to the sudden turn that came on this summer. Sensing that consumer demand was weakening, starting with PCs and smartphones, many hardware makers took steps to reduce their bloated inventory, halting new orders and sending chipmakers into a tailspin. Other factors have contributed to the oversupply, including a jump in global chip manufacturing capacity. According to Dan Hutcheson, chair of chip research firm VLSI, global wafer capacity — the silicon discs on to which chips are etched — climbed from 1.06bn square inches early last year to 1.22bn this September, well ahead of normal industry expansion rates. The overcapacity may well become a permanent feature of the market, said Pat Moorhead, an analyst at Moor Insights & Strategy. Attempts by the US and EU to reduce their reliance on global supply lines is leading to a “balkanised” chip world, said Moorhead. With each large country or region seeking to build in enough surplus capacity to cushion itself against unexpected shocks, the overcapacity could become structural, he added. But an end to the first stage of the sharp downturn could at least be coming into sight. Qualcomm last week predicted it would take smartphone makers two quarters to burn off their excess inventory, with the low point in its sales coming in the current quarter. AMD also noted a potential clearing of the excess by next Spring, pointing to a period of greater stability. Some analysts also said that the surprising severity of some of the latest forecasts, which were well below Wall Street expectations, suggested a bottom in the market may be near. “We’re getting to the capitulation point,” said Hutcheson, referring to an apparent willingness by chip companies to throw in the towel and write off some of their own excess inventory. An end to the sharp correction, though, would still leave the industry vulnerable to any broader weakening in the global economy, with a hoped-for recovery in chip demand in the second half of next year at stake. Global chip sales could fall anywhere from 6-20 per cent for the year as a whole, according to a VLSI forecast — though as the surprising severity of this year’s slump has shown, even predictions with such wide ranges can miss the mark. Demand is still red-hot in some chip markets, according to executives at companies that have reported earnings recently. Car companies top the list, as supply shortages continue and demand rises for electric vehicles and increasingly sophisticated driver-assistance systems. The latest spurt in capital spending by the biggest tech companies has also ensured strong sales of chips for large-scale data centres. Shortages of analog chips — which are used in things like power supplies and sensors — have also continued. Yet some of the biggest markets for semiconductors are entering a profoundly different period. After booming during the pandemic as millions of people were forced to work and study from home, PCs sales have fallen off quickly, with many forecasts pointing to a 20 per cent decline in sales this year, to around 275mn. Sharp disagreement over the outlook for next year has highlighted the potential for further disappointment. Intel said it believed demand for PCs had moved permanently higher as a result of the pandemic, and last month issued a surprisingly robust forecast for sales of 270-295mn. Rival AMD, by contrast, predicted another 10 per cent decline — suggesting that sales could fall to around 250-255mn and taking the PC industry back to pre-pandemic levels. The latest round of quarterly earnings reports also revealed that the weakness first seen in consumer demand for PCs and smartphones had spread, with the gaming, enterprise and industrial markets all reported to have seen a softening in order growth. At the same time, Texas Instruments, which sells into a broader range of markets than most chip companies, predicted it would see a weakening in demand from all of its end-markets before the end of the year, with the exception of carmakers.

This today from Ben Thompson Stratechery newsletter:

Mark Liu in Time

Let me start off with a fun bit of cultural trivia: there was a profile earlier this week of TSMC, including access to fabs, employees, and an exclusive interview with Chairman Mark Liu. What is perhaps surprising is that the profile appeared in TIME, which is not exactly well-known for its tech coverage, and seems years past its prime in terms of groundbreaking reporting generally. TIME, however, is a big deal in Taiwan — everyone here recognizes and respects it — and it’s the go-to publication for politicians like President Tsai Ing-wen; I suspect TSMC had a similar calculus.

Anyhow, the most interesting tidbits in the article were from Liu; they are scattered throughout the article, so I’m going to just jump around.

With regards to the chip shortage:

The semiconductor-chip shortage first got boardrooms sweating around February, when average order-to-delivery times for chips stretched to an unprecedented 15 weeks because of a confluence of factors: a pandemic-induced economic slump prompted carmakers to prematurely slash chip orders, which soon rebounded as chips were hoarded by firms that feared being embroiled in the U.S.-China trade and technology war. Amid what was described as a global chip shortage, more chips were being sent to factories than were leaving them in products, meaning “there are people definitely accumulating chips who-knows-where in the supply chain,” says Liu.

To fix things, Liu ordered his team to triangulate different data points to decipher which customers were truly in need and which were stockpiling. “We are learning too, because we didn’t have to do this before,” says Liu. It forced him into tough decisions to delay orders for valued clients whose immediate need was judged to be less acute. “Sometimes [customers] may not be satisfied, but we just have to do what’s best for the industry.”

Jalopnik had a good piece about the fundamental mismatch between chipmakers like TSMC and car companies: the former only want to build fabs on newer nodes, which produce more chips per wafer, while the latter have already qualified chips used on older nodes, and don’t want to have to re-design and re-certify new chips on newer nodes. In the long run it is the car companies that will have to give in: there simply is no good economic reason for TSMC or anyone else to build new “old” fabs. The certainty that no new supply is coming online, though, unsurprisingly has led to hoarding, and it has fallen on TSMC to backtrack a bit on its “the customer is always right” philosophy and pick who gets new chips first.

On TSMC and Apple:

TSMC’s more recent run of success has been linked to one client in particular: Apple. The Cupertino behemoth outsourced the manufacturing of its chips to Samsung for the first six generations of iPhones. But after Samsung launched its own competing Galaxy smartphones, Apple in 2011 brought a lawsuit over IP theft, which was ultimately settled with an award of $539 million to the American firm. That dispute was a boon to TSMC as Apple sought to extricate its supply chains from Samsung and avoid any partnerships that could burnish a potential rival. It was reassuring that TSMC was a dedicated foundry business that wouldn’t stray from its lane. Apple remains TSMC’s biggest client today. “It is a trust business,” says Liu. “We do not compete with our customers.”

However, Apple insisted that it wanted a new node for each iteration of iPhone. Because Apple prides itself on never missing a launch of its crown jewel, TSMC was under enormous pressure to come up with constant advances. So instead of combining lots of new technologies to double power every two years, it pioneered small advances annually. “People made fun of TSMC, saying, ‘Oh, that’s not a real node,’” says Nenni. “But taking these baby steps helped them learn these new technologies. And they laughed all the way to the bank.”

This underestimates Apple’s involvement: the iPhone company also puts up a huge amount of money to help push TSMC to new nodes, and in exchange gets the newest chips first. That did, by the way, happen this year as well: the A15 is, like the A14, a 5nm chip, but it is on the second generation (TSMC calls it N5P), which affords 7% more speed at the same power, or 15% more power efficiency at the same speed (Apple ultimately increased the clock speed but more than made up for it with architectural improvements in efficiency).

The bit about TSMC not competing with its customers, meanwhile, has always been a critical part of TSMC’s success, and will always be a challenge for both Samsung and especially Intel.

TSMC and the U.S.

Finally the U.S.:

Despite America’s dominance of chip design, its lack of manufacturing capability remains worrisome for policymakers, who are trying to bring more fabs onshore. While remaining a TSMC customer, Intel is revamping its foundry business, building two new fabs in Arizona at a cost of $20 billion. Last year, TSMC committed to building a $12 billion fab, also in the Grand Canyon State. It is also exploring more plants in mainland China, Japan and Europe.

Liu is candid about the reasons for that U.S. investment and its limitations. It was prompted by “political nudges on our customers,” he says, insisting that “semiconductor localization will not increase supply-chain resilience.” He says it may even “degrade resilience”…

What would make more sense, says Liu, is for the U.S. to ensure it stands at the frontier of the next great advance. Its lifesaving prowess in mRNA COVID-19 vaccines, for example, exists only because of huge investments in genomics and biotechnology over the past 40 years. Instead of futilely chasing and localizing aspects of the semiconductor supply chain, Liu suggests plowing that same money into developing the next great leap. “The U.S. should focus on their strengths: system design, AI, quantum computing, those forward-looking things,” says Liu.

Liu, in a vacuum, is obviously right. TSMC running globally distributed fabs is less ideal than having all of its engineers and equipment in the same place. And of course TSMC would rather see the U.S. government spend money on something other than a TSMC competitor, not that it will make a big difference:

The crisis sharpened focus on access to technology that the U.S. invented and still designs better than anyone else but doesn’t manufacture at scale anymore. Biden’s $2 trillion plan to fix American infrastructure included $50 billion to boost semiconductor competitiveness. That was on top of $52 billion committed under the U.S. Innovation and Competition Act, which was passed by the Senate in June and is aimed squarely at competing with China in all areas of technology. Yet TSMC alone is investing $100 billion in new capacity over the next three years. It’s a staggering sum, though for Liu, “the more I look at it, it’s not going to be enough.”

This is the core problem: there is just no obvious reason for any company to not work with TSMC; if the U.S. wants to change that reality its investment will have to be an order-of-magnitude larger. That’s what South Korea is doing, pledging to spend $450 billion to expand its chip industry, which, given Samsung’s long experience as a foundry, is already a much more likely alternative than Intel’s fledgling efforts (GlobalFoundries, which filed for an IPO, doesn’t have cutting edge foundries; they do have a customer service org and a decade of experience as a foundry, though, which is why Intel should buy them).

That leaves one other reason for Liu to encourage the U.S. to spend its money elsewhere: TSMC, which is still partially owned by the Taiwanese government, is less a bargaining chip than it is an indispensable asset, making the likelihood that the U.S. would come to Taiwan’s defense, if necessary, that much more likely.

This from this week's Economist.   https://www.economist.com/business/intels-turnaround-and-the-future-of-chipmaking/21804288?utm_campaign=the-economist-today&utm_medium=newsletter&utm_source=salesforce-marketing-cloud&utm_term=2021-09-07&utm_content=article-link-2&etear=nl_today_2

A few highlights:

"Intel, wants to throw the windows open. The American semiconductor giant has long guarded its core chipmaking business as jealously as Microsoft did its OS. After years of product delays, misplaced technology bets and changing management, it is ready for some fresh air. “Our processes, our manufacturing, our intellectual property through our foundry services [producing processors for other chipmakers]: all will now be available to the world,” professes Pat Gelsinger (pictured), Intel’s newish boss."

"f successful, Mr Gelsinger’s strategy could reshape a $600bn industry at the heart of the fast-digitising global economy for the better. Failure could, in the short run, compound the chip shortages that are making life difficult for manufacturers of everything from cars to data centres. In the longer term, it could lead to further concentration of the already cosy chipmaking market, with Intel increasingly eclipsed by rivals. And it may cement Asia’s dominance of the industry, creating all kinds of geopolitical complications."

"As the tech industry has grown bigger, more diverse and more networked, this once dominant “integrated device manufacturer” (IDM) model has fallen out of favour (just as vertical integration became a drag for Microsoft as other tech “ecosystems” popped up). As with the Microsoft of old, Intel’s arrogance and insularity discouraged other chipmakers from working with it, for instance by combining chip designs. Instead they ploughed their own furrows, focusing increasingly either on designing chips (for example, AMD, Arm, Nvidia and Qualcomm) or fabricating them (notably Taiwan Semiconductor Manufacturing Company, TSMC)."

"Now, though, the company is being overwhelmed by open systems like that of Arm, whose blueprints are used in most of the world’s smartphones (a market which Intel missed) and are starting to appear in data centres—and which was last year acquired by Nvidia for $40bn (though trustbusters may yet scupper the deal). At the same time TSMC took advantage of Intel’s technological and management missteps to pull ahead in both cutting-edge technology and production volume. Both TSMC and Nvidia are now worth more than twice as much as Intel (see chart), despite lower revenues and profits."

"Cutting-edge chipmaking involves around 700 processing steps and many nanoscopic layers printed and etched on top of each other. Adding to the complexity, Intel will at last fully embrace “extreme ultraviolet lithography” (which TSMC and others have already been using to great effect). The company’s announcement in late June that it would postpone production of next-generation server processors for a few months hints at the trickiness of the task."

The current chip shortage is well documented- in particular the effect on car manufacturers. The first order conclusion would be that current manufacturers are onto a winner and hence investing in established chip manufacturers would be a no-brainer. 
There are, however, other second order factors at play, which may mean that this might not be the smartest plan. 
Firstly, many companies have started to develop in-house manufacturing of their own chips. Apple has ditched Intel, AWS is starting to design their own chips, Tesla is also looking to de-couple from established chip design and manufacture. 
Secondly, is the phenomenon of semiconductor  nationalism. It is no secret that Taiwan is in something of a bind trying to placate increasingly onerous US conditions on chips that are created using US patents that might be on-sold to Chinese companies. Yet, China is its largest market. China, the Us and the EU have stated an intention to become self-sufficient in semiconductor supply as they are not only essential for economic prosperity but are also integral in numerous modern military applications (fighter jets, advanced warning systems, missiles etc)  Taiwan's independence and security from a mainland Chinese invasion is no longer as impossible as it once seemed  

Thirdly, the cure for high prices.....is high prices. Other entrants will seek to take advantage of high margins by entering the market. As often happens there is then a flood of new products with a collapse in margins and profitability. 
Lastly, is the exorbitant cost of building a new fab- it was recently a couple of billion. Taiwan semiconductors is planning a new 2nm chip manufacturing facility at a cost of 19.5 billion dollars. Returns on investment are not guaranteed if numerous other entrants are also doing the same. 
 

I think it's a difficult market to work out how it will all play out and have consequently not explored much further. 
that whole thought bubble didn't even include Brainchip!!