8 Comments

Very interesting stuff about chiplets. What are the downsides of them? I would assume they run into interconnectivity issues. I wonder how their efficiency compares at different sizes.

Expand full comment

Thank you. One challenge would be getting different vendors to manufacture components that are compatible with each other, making it easier to package chips as chiplets. Which is why it's particularly notable that China has already implemented two sets of industry standards for chiplets as of last year.

Expand full comment

Leading-edge and trailing-edge engineers could be described as hot competitors, right now, for years, in batteries.

"Leading edge" would surely be a new breakthrough battery chemistry and design that throws Lithium-Ion into the shade, where "Nickel-Cadmium" now dwells - haven't seen a "Ni-CAD" in years.

"Trailing edge" would just be picking away at the efficiency of manufacture of the current battery, now >30 years old.

And, for years now, Li-Ion has been getting cheaper, faster than they can ramp up leading-edge would-be replacements into its competition space.

Trailing edge is undoubtedly more of the economy, for instance:

Internal combustion engines got 30% more efficient, the last 50 years since the 1973 oil crisis.

The improvement was obscured by people taking that brilliance, and using it to buy giant SUV and Trucks, rather than save gas...but arguably it created the new product category; an SUV would be much more gas-expensive and have lower sales.

In my own job, PVC plastic for water pipe didn't change the industry or anything, but it costs half what ductile iron did, can be moved around in the trench by one guy instead of needing a backhoe, halving construction effort ... and it NEVER corrodes, eliminated literally millions of main breaks over 40 years now.

Just one of a thousand things that get better every year with almost nobody outside the given industry, even knowing about it.

Expand full comment

This is a very useful article. Over the past few decades, America has focused far too much on software, and very little on advancing industry. I believe it is partly a result of a faulty mindset, and have outlined this idea here: https://swiftenterprises.substack.com/p/digital-involution

Expand full comment

Chiplets are neat and illustrate the problem with the US’s attempt to suppress Chinese chip innovation to try and stay “ahead.”

The US would be better served uplifting itself; improving its education system, reforming the immigration system to gather more talent, and investing more in science and technology.

Expand full comment

"latecomers like China working to catch up with forerunners like the US?"

Several years ago China overtook the US in every field of scientific research and in institutional capacity: 7 of the top 10 research institutes are in China, one is in the US.

The US retains the lead in 1980s patents

Expand full comment

Thanks for reading and engaging. Chinese-language sources, including those from the government, do often refer to China as being in a latecomer/second-mover position relative to the US.

Expand full comment

Indeed, I get frustrated whenever I hear people trot out the old "China copied it" trope. China is the second largest spender in R&D, graduates far more STEM PhDs than the US, and has led the world in patent applications and scientific journal publications for years.

Expand full comment