Last week NVIDIA CEO Jensen Huang talked about how NVIDIA wants to be the TSMC for AI in 2014 GTC keynote. Which reminded me to update my 2021 wish for a TSMC for smarter cars. Despite recent cooling down of the EV rush, I still bet on the same strategy, just for a bit longer run. Because increasing geopolitical concerns and desires for shorter supply chains & local manufacturing, the window of opportunity to redefine the Smart Car industry and accelerate new technology adoption is still wide open. Here is the updated thesis. Let’s let the bullets fly a bit longer 😉.
Lower the entry barriers + the scaling law
With a smart car foundry (design to deliver as a Service), small OEMs may be able to compete with big OEMs on price, performance, quality and even user experience. Similar to the chip industry, which used to be a business only for the big guys, e.g. Intel. Because only they could afford the huge upfront investment on developing the tech & building the fabs to make chips. After TSMC started to provide the foundry service (chip manufacturing as a service), not only NVIDIA, AMD, Qualcomm and many small chip design houses can play & complete too. But also, it enables the new players to build your own chips. e.g. Apple’s A chips for iPhones & M chips for iPads/MacBooks, and Tesla’s chips for Self Driving & Dojo server farm. By 2021, Intel finally woke up & has been trying hard to catch up by Intel foundry service since.
A smart car foundry & its ecosystem
A smart car foundry will play a critical role to being the economies of scale to its ecosystem, e.g.:
- The Build-To-Order model can empower the OEMs to pick & choose the dream cars for their brands & target market segments with the state of the art technologies, such as platform chassis, 48-volt electrical system, Software-Defined Vehicle, Smart In-Vehicle-Infotainment systems, event Self-Driving services & more.
- Product design & development are from top talents augmented by the bleeding edge technologies, such as designing together in Omniverse Cloud streams to everyone’s reality by Vision Pro. For example, the iteration cycle will be much shorter without a $650K of clay model nor even more expensive concept cars in the loop.
- Cars get “smarter” over time with new features, improvement & security enhancement from an app store & Over The Air updates for better user experience with less risk & cost of physical recalls.
Job creation at scale
Auto is a top industry contributing to GDP & job market. But, it had always been the game for big auto nations such as Germany, Japan, Mexico/US. From 2009 to 2022, China poured over $29B into EV subsidies and tax breaks according to a MIT Technology Review report. As a result, Chinese car makers become rising stars. In 2023, China is estimated to be the №1 as world’s auto exporter for the 1st time, shipping 5M cars & bring in $102B in value. Whereas, the №2 Japan only exported 4M cars as Reuters reported. To be fair, Japan & South Korea followed a similar play book to pave the way of industrialization.
The white label knight
Last week, Intel announced $8.5B in direct funding, up to 25% tex credit on more than $100B investments & eligible of $11B loans from Uncle Sam to be fabs in US. Not every county may afford to build a fab, but many may dream for a car factory/a lot of jobs. Why? Psst! in the US, the auto industry directly employs over 1.7M people.
So, how much will a nation invest to bootstrap the mass job creation, and therefore the economy? How valuable is it for a white knight can make local kings of car brand, factory, changing network & service franchise? Let me know if you want to know more how that can be possible by [x] Platform as a Service model.
