Back in the 80s it was the space race between the US and USSR, who could get first to the moon. The biggest and most powerful engines to take a human to the moon. Now is the complete opposite with China, the microchip race is who can design the smallest engine (chip) to power the machines on the datacenters and their AI. On this video, the race is going into a different lap and the stakes are and gaps are closing in as China has make some advances on the missing tech they need to produce the same type of processors as the US.
Last year something similar happened when Deepseek ruin the party of the US based AI giants like Google, OpenAI and Meta. When a Chinese startup was able to compete at a 10th of the cost. The market froze and a pile of bricks sent the plans of these companies down to the dark haze of uncertainty, fear and doubt.
This report also goes about the geopolitical implications of what China could get if they get to compete with the American counterparts and how can impact the global economies. The story of ASML and how companies that supply them; like Carl Zeiss AG, could loose the technological advantage.
ASML’s high valuation moat does not exist in isolation; it depends on a hyper-specialized global ecosystem. For instance, ASML relies on Germany’s Carl Zeiss AG for the world's most precise optical mirrors and lenses required to guide extreme light wavelengths. Export controls orchestrated by the US, the Netherlands, and Japan restricted whole machines, but loopholes initially allowed foundational components, optics, and sub-assemblies to flow into China. Chinese firms leveraged these gaps, acquiring DUV equipment to reverse-engineer and domesticate core sub-systems. However, replicating the full EUV supply chain—consisting of over 100,000 hyper-precise components—remains one of the most complex industrial tasks in human history.
However the core conclusion is that China still a relatively long way to go until they can really be competitive. And even if these news put the race in a relative choke-hold, the exaggeration and hyperbolic are still overrated to the actual data that has been proven. China’s current DUV milestone is a crucial proof-of-concept for its domestic ecosystem, but it is merely the first step on a long runway. Until China can mass-produce EUV systems with high yields, Western foundries and equipment leaders maintain a substantial technological lead in efficiency, cost, and raw computational power.
Is an interesting race and like most technologies, a breakthrough might make this race obsolete faster than the race will take to finish. However, what are your thoughts on this? Are you interested in this at all? You still have that old way of thinking that AI is consuming all the water in the world? Or you still think DeepSeek is the doom of all the AI industry? Let me know on the comments and hopefully we can get some more updates on this situation.
