I choose to follow this up because AI and healthcare have become dominant discourses. The basic fundamental building block is “chips” that come in various forms, sizes and now specialist chips for AI. The big tech has moved into designing their own custom chips and getting them manufactured with TSMC, Taiwan. I am linking to two fascinating historical insights. One from Japan when the rivalry with the US was fierce, and another from the UK, which hasn’t been able to capitalise on the value of its academia-linked research (ARM).
First the Japanese story. These are critical to be understood for those countries embarking on their own fabs. Historical perspectives help avoid the pitfalls that others suffered.
ETL began with a simple transistor study group led by Watanabe and other ETL researchers. The ETL director flew to the United States and brought back transistor samples. The GHQ had set up libraries in Japan and those libraries had subscriptions to Bell technical journals. The researchers read them. By 1951, Japanese scientists at NTT and Nippon Electric Company or NEC succeeded in creating proto-versions of the transistor. However, this remains far behind what the rest of the industry had to offer.
Rest of the fascinating newsletter offers multiple perspectives on the rise and fall of chip design and manufacturing, and how South Korea copied the Japanese playbook to exert dominance in DRAM.
To modify a Japanese saying: America pounded the rice, Japan knead the dough, and South Korea ate the DRAM cake. Japan’s reign as king would not last for very long.
Japanese are trying to build a fab to address the mature nodes (automobiles and electronics) and will possibly refrain from the “cutting edge processes” for the time being. However, chip design and manufacturing are taking geopolitical overtones, as it represents the next wave of technological race. If you extrapolate the whole story and take a zoomed out view, Europe, India, China and the US are leaping into the unknown. It is important to focus on supply chains and ensure no breakdown occurs if “cloud-as-a-service” or “on-premises” computing systems become paramount for the country’s GDP growth.
A more humbling story from the UK, published in the Financial Times:
The small number of British tech success stories is a longstanding disappointment to policymakers. Countless reports over many decades contrast the UK’s poor record at commercialising breakthroughs with its scientific prowess. This is demonstrated by the high rankings of its universities and its success at winning more science Nobel Prizes than any country other than the US. Failure to capitalise on the first large-scale electronic computer (built to decipher German codes in 1944) is the first of many examples. Unicorns — venture capital-backed start-ups worth $1bn or more — are hailed as a sign of the UK’s success in nurturing tech businesses. But their high valuations typically reflect investors’ expectations that they will prove attractive to buyers. Companies with important innovations generally have more value to an acquirer than to a long-term financial investor. Takeovers are the norm unless the management retains control.
Bulk of the write up dovetails the UK’s “research story” around failure to commercialise, but ARM (ironically owned by Japanese) has been declared “critical” and fended off NVidia’s buyout. It’s still on the selling block and is possibly a drag on SoftBank.
ARM doesn’t manufacture-it designs chips. Yet, it is the only crown jewel that the UK has been left with. After they stole the Kohinoor (and trillions from their colonies), they are running out of options and dangling their “passport” to “attract and retain” talent (in the face of tottering NHS and inflation). Let’s see what genie they pull out of their hats.