Foreign media in-depth analysis: how did the UK quickly slide from silicon hegemony?

Jiwei network news, foreign media thechipletter recently released a long article saying that in the short period of the 40s and 50s of the 20th century, silicon chips seem likely not to be invented on the east coast of the United States, but in the United Kingdom. And deeply dissected how Britain quickly slipped from silicon supremacy?

The following is the compilation of the set micronet:

Of all the inventions of the twentieth century, one of the most important and certainly the most symbolic was the transistor.

Just before Christmas 1947, John Bardeen, Walter Brattain and William Shockley first patented the solid-state switch at Bell Labs in New Jersey, widely regarded as the cornerstone of the computer age. As Chris Miller chronicled in his brilliant book The Chip Wars, semiconductors are now at the heart of almost everything we do – "most of the world's GDP is produced by semiconductor devices" – so the advent of the first transistors represented a huge turning point.

But there is a little-known story that during the brief period of the 40s and 50s of the 20th century, it seems likely that silicon chips were invented not on the east coast of the United States, but in the United Kingdom.

Much of the early research in semiconductors (the term technically refers to a whole class of materials between conductors and insulators) was mostly related to the invention of radar. Silicon and germanium (the first semiconductors were actually made from germanium) are considered materials for radar detectors, while cutting-edge research into radar and semiconductor rectifiers was conducted in the UK.

As a result, much of the early work at Bell Labs, MIT, and Purdue University in the United States was based on the discovery of these strange and exciting materials in the UK. If Bardeen, Bratton and Shockley hadn't invented the transistor at that time, it would most likely have been invented by someone in England (or for that matter, equally advanced in French research). In fact, it is said that within a week of Bell's final announcement, the transistors were actually made in the UK. According to the history of semiconductors, "the time for this idea has indeed arrived".

Even in the 50s of the 20th century, the UK was still at the forefront of semiconductor research. The Royal Radar Agency (RRE) in Malvern was a hotbed for thinking about this new technology, which at the time was still largely in the prototype phase. Yes, transistors have been invented, and they seem to be a replacement for glass vacuum tubes, which have previously been used as switches and amplifiers in primitive computers and radios. But the real breakthrough, the one that helped give birth to the computer age, is still a long way off.

The idea is to integrate multiple transistors on a single circuit board, and integrated circuits are arguably a bigger breakthrough than the transistors themselves, because this is where what we know as "silicon chips" come from.

Traditionally, integrated circuits were invented by Jack Kilby of Texas Instruments and Robert Noyce of Fairchild Semiconductor in the late 50s of the 20th century. Kilby eventually won the Nobel Prize for the invention. Noyce died before winning the award, but he died before he founded Intel and became very wealthy.

But the big idea was actually formalized five years ago, not by the Americans but by Geoffrey Dummer, an Englishman from Hull. Dahmer, who worked at RRE, is now almost forgotten, but he is one step away from changing history. In 1952, he laid out the next leap forward in semiconductors in a speech in Washington. While most people were still figuring out how to use simple transistors, he envisioned a single semiconductor substrate consisting of the "insulating, conductive, rectified, and amplifying" parts of the transistor, "directly connected to electrical functions through areas in the cutting layer." This is integrated circuits.

It's not just an idea. Dahmer began experimenting with making the world's first integrated circuit. He came very close, and a few years later produced something that almost did the job, and he showed a silicon block with four transistors at the RRE event in 1957. The problem is, this silicon block is not practical. This is just a model. Manufacturing silicon chips turned out to be much more difficult than thought.

Over the years, Doomer and RRE (and Plessey, an electronics company) have tried to turn slices of silicon crystals into integrated circuits, but have always encountered physical difficulties. There are too many impurities in the material, which means that semiconductors simply cannot work. They tried several ways to connect components to the circuit, first using a "thin-film" process and then a diffusion process. The years passed, but Dumo continued to produce failed works.

What they encountered was a lesson that I explored in World of Materials. These amazing installations are not just a manifestation of mental work and ingenious design, but the embodiment of the world's most advanced materials science. Dummer, RRE, and Plessey hit the limits of physics and chemistry in the '50s. In fact, for most of the early days of semiconductors, one of the biggest challenges wasn't a lack of brainpower, but our inability to create something pure enough to be converted into a semiconductor, and the misalignment of one of the atoms could ruin a device. This is especially difficult for silicon with extremely high melting points.

Even today, semiconductors are still an alchemy — humans have the ability to convert a fairly common material (quartzite mined from quarries) into silicon wafers, which in turn are more valuable than gold into computer chips. It's a magical universe that's mostly hidden from our sight, but we all depend on it. One of the most exciting chapters in the book tells a story I've never read anywhere else: a grain of silicon's journey from leaving Earth's quarries to becoming a chip in your device.

I believe this is the most extraordinary journey in the modern world — and perhaps most interestingly, much of this journey takes place before chips enter semiconductor manufacturing plants in Taiwan and South Korea, where they are "manufactured." Our biggest concern these days is just the tip of the iceberg. In Material World, I also try to explain what happens underwater.

Part of the reason why integrated circuits were invented and mass-produced in the United States and not in the United Kingdom is that they do a good job with materials. Kilby and Neuss, and perhaps more importantly, Jean Hoerni, who proposed the planar process still used in chip manufacturing today, are using materials science to usher in the semiconductor age.

But in the '50s, there was a tantalizing possibility that modern, groundbreaking technology might have been invented and prototyped in rolling green English countryside rather than in California or Texas. In fact, today, Doomer has largely been forgotten. RRE was sold to defense company Qinetiq, and Malvern's site was sold a few years ago — it sounds like it's being turned into a residential area, though some intrepid explorers snapped some photos before demolishing it years ago. Plessy was sold to Siemens, which later expanded its semiconductor business into part of Infineon, which is now one of Europe's largest semiconductor companies. The UK's semiconductor industry has never reached the heights it once promised.

This brings us to the Integrated Semiconductor Strategy released today by the government. It's a strange, thin document (literally: despite years of work, it's only 40 pages, plus a summary). The document does not pretend that Britain can compete with giants in the semiconductor field such as Intel or Taiwan's TSMC. The amount of money promised to the UK sector is insignificant compared to the money thrown by the US and the EU. As someone has said, the total 10-year commitment in the UK is about the capital expenditure that a company like TSMC invests every two weeks.

It's easy to squint and say: well, things could have been very different, but it's also worth noting that the UK has tried to join the race many times before and all failed. A few decades ago, the UK had centres of excellence in certain disciplines (including compound semiconductors, which even today are held as our last hope). In other words, we've been here before.

All of this might lead you to conclude that the government should take a step back altogether. But it's also worth noting that without government support, no chip giant in the world today would exist — sometimes with simple grants, sometimes with trade protection, sometimes by providing a reliable large customer (such as the military and space programs, which helped fund much of America's early chip manufacturing). Like it or not, the industry often relies on government help — at least to get it off the ground.

In other words, despite the UK government's allergy to industrial strategy – I've heard that the term is actually banned in the UK – it currently faces a range of problems that it would be difficult to solve without an industrial strategy: from hydrogen to batteries to semiconductors to steel.

What's more, the geopolitical environment it currently faces is one in which everyone else, including the UK's main ally, is wholeheartedly pursuing an industrial strategy, which means it needs to decide whether to do the same, whether to join forces with neighbors or allies (Brexit doesn't seem to help in this regard), or whether to accept laissez-faire policies and watch much of its tech industry leave these coastlines.

But the problem with industrial strategy is that you always end up having to put money behind the back of ideas and businesses that sometimes don't work out. This is politically embarrassing. You just don't know which ones will work out and which ones won't.

I have a little personal experience in this area. Last year, I did a story about the Newport fab, one of the UK's last major remaining semiconductor manufacturers. One of the companies in that story seemed to have hope for success, and it was called Rockley Photonics. There's a lot of hype around Rockley. It is making an unusual semiconductor that is rumored to end up being a health sensor in future versions of the Apple Watch, allowing users to check their glucose levels noninvasively.

So I introduced Rockley in that article as an example of a UK company that might have the opportunity to go bigger. We interviewed the founder in a TV report. A few months later, almost sudden news revealed that Apple had abandoned Rockley as their photonics supplier, opting instead to work with a division of TSMC to produce the chips. Rockley filed for bankruptcy in January.

If the government gives it more support, might Rockley have a chance? This is difficult to determine, but it is a long-term problem with industrial strategy. You put taxpayers' money at risk. Sometimes these bets pay off, but often they don't.

Consider several other examples of success/failure in this industry.

The first example was the wafer machine, which was seen as the next big thing in semiconductor design for a period in the late 70s and early 80s of the 20th century. The UK government sees this as the future, and a future in which it can be reasonably contested. If the chipset, a parallel chip, performs well, it could turn the UK into a chip-making powerhouse. As a result, millions of pounds of public money were poured into a Bristol company called Inmos, which has a manufacturing facility in Newport. As the industrial strategy evolves, this seems like a smart move.

Around the same time, in the early 80s of the 20th century, the Dutch electronics company Philips joined forces with a company called Advanced Semiconductor Materials International to create a new company called ASM Lithography (ASML). The idea is to make large, complex machines for "burning" transistors onto silicon chips (more on this in Materials World). This is a bold move, especially since the market for lithography is completely dominated by Japan. The first few years of ASML were very tough. The company came close to collapse, but received help from Philips and generous support from the Dutch government.

A few years later, in the late 80s of the 20th century, Morris Chang was invited to create a semiconductor company in Taiwan. Zhang has spent most of his career in the United States, working at some of the earliest semiconductor companies, including Texas Instruments. At the time, most semiconductor companies tended to produce chips of their own design, but Zhang chose to do something else: make chips on behalf of other chip designers. The birth of TSMC was difficult, but it also has broad support from the Taiwanese government, which has given it considerable tax breaks and attracted wealthy Taiwanese investors to capitalize it.

You probably know what happened next, at least a few of these examples. TSMC (remember, it was founded around a decade ago in the UK that it shouldn't be in mainstream chip manufacturing, because the US has planted it all) is now the world's largest semiconductor manufacturer. Today, ASML is the world's only machine builder, and TSMC and other companies use it to make the most advanced logic chips — the ones found in today's smartphones. However, it is important to note that for a long time, the conventional wisdom was that these businesses had no hope. They look like dumb.

On the other hand, the transport aircraft began to be hyped, but certainly a failure. In fact, it may have been a loser; Britain chose the wrong technology. This happens quite often (quite a lot, actually).

However, after the Labour government began to implement this semiconductor strategy in the late 70s of the 20th century, the Conservative government abandoned this strategy and most of its support for the industry in the 80s of the 20th century, which did not help the chip machine either. Inmos was soon sold to a French-Italian company and became part of STMicroelectronics. Companies like Plessey and GEC, once side-by-side with their European counterparts, have now shrunk, look vulnerable and are being sold. Most of the intellectual property was transferred across the strait to France, Germany and Italy. Today, the UK's main heritage is a cluster of semiconductor design companies located in the Bristol and Newport fabs.

Inside the Newport fab

But on the other side of the UK, there is an undisputed success story. Acorn, a Cambridge-based microcomputer company, makes some computers but struggles to gain a foothold in the field. However, Apple found that the chip architecture Acorn uses in its computers, Reduced Instruction Set Computing (RISC), can be deployed in its Newton Notepad. This gave rise to a new company, Advanced RISC Machines or ARM.

Interestingly, ARM's success is not so much the product of the government's industrial strategy as it is the ingenious strategy of its directors. Just as TSMC chose not to make its own chips and make anyone else's chips, ARM chose not to make its own chip designs and licensed its architecture to other companies. Today, ARM architecture is completely dominant in much of the semiconductor world. It is a true global champion.

ARM was acquired by Japanese investment group SoftBank in 2016. In any other country, it's hard to imagine that an acquisition of such a large enterprise, especially one whose architecture is used for security-sensitive devices, would be approved. In fact, the British government kept "golden shares" in certain important defense companies, including Rolls-Royce and British Aerospace, precisely to prevent such things from happening. However, the acquisition came shortly after the Brexit referendum. Theresa May saw it as a vote of confidence in Britain and allowed it to pass. Britain's last semiconductor star is in the hands of foreigners. Now, SoftBank is getting ARM back on the market, not in the U.K., but in the U.S.

Today, while ARM still has an important presence in the UK, some wonder how long its connection in the UK will last. But the company is important because it is the UK's only real foothold in the global semiconductor market. This area is also documented by materials scientist Richard Jones, who believes that the UK government has a good reason to buy ARM, bring in gold stocks, and then sell it again.

Such an idea is unlikely to be pursued in a Conservative government. But what's interesting about this moment is that old assumptions about what is acceptable and what is not are rapidly shifting. The UK government may not like to use the term, but today's Semiconductor Strategy document represents the first glimmer of light that could finally become a true 21st century industrial strategy.

This can be a somewhat sad, unimaginative start. The numbers are small and the ambition is weak. There is no mention of energy costs – one of the largest inputs to semiconductor manufacturing and the biggest obstacle to the country's heavy industry. The document seems to completely ignore the fact that the UK has a large chemicals industry and therefore has the potential to compete in the manufacture of some of the chemical inputs used in chip factories around the world. But hey, this is a start.

It's a sign that this administration may be ready to start wrestling with today's world, where countries are vying for supremacy in silicon as well as lithium, copper, hydrogen, and more. It may cost a lot of money and a lot of money will be wasted, but the result of all this is likely to be a new industrial revolution, not to mention a cleaner, healthier world. Britain may enter the fray later, but this may be a sign that mentality is shifting. There may be more such shifts in the future.

Pre-shop the world of quality

I also hope this is also the beginning of something else: the world we live in today without dematerialized epiphany. It relies on very physical, very real products that are really hard to manufacture (as Geoffrey Dummer will tell you). It depends not only on our ability to turn good ideas into clever programs that run on microprocessors, but also on our ability to turn raw materials dug out of the ground into ultrapure silicon wafers, and into the gases and liquids we use to make them.

Material World attempts to tell this magical story from the bottom up. It won't help you develop an industrial strategy for semiconductors, steel, or any other industry. It won't help you choose between chipsets and TSMC in this world. However, it may help you understand some of the hidden economic underlayers that we have long taken for granted.