>We always were sort of maintaining two different bills of materials of Chinese componentry and Western componentry because they're different. Then we produced five different iterations of the Librem 5 phone through Chinese contract manufacturing. And we iterated through those five changes over the course of about 18 months. At that point, we finally had a production ready product. And then we were able to take everything that we did and bring it to US soil.
How many iterations can they get done in US in 18 months? That's what's going to kill Made in USA, if you can't get the design done on time, not a few $100 extra in PRC vs west sourced BOM, but millions more spent on development over longer time frame because lack of talent. Is the short/medium term solution still to send "homework" to Chinese prototype teams? I suppose economics of PRC speed and few prototypes > 100%+ tariffs.
Engineers in the US would be able to get five iterations done using US fabs and assembly too, it’d just cost a ton more to get the same lead times and pay for the NRE.
Last time I worked on something of this complexity in 2019, 1 week turn around prototypes would cost $2-5k for the assembled PCBs from China but $30-50k in the US. It also took a bit more effort and inventory to make sure all the parts were stocked or shipped on time, which is a problem when you can’t visit Shenzhen’s malls if you’re missing a part. Once the first iteration was done, we were averaging between one and two months per revision. It’s very doable but nobody except medical, defense, and aerospace are willing to pay the price.
Kind of a pointless debate point (and economic plan) as it will always come down to humans and market conditions. Where the humans are who specialize and have deep experience in that stuff chose to work. You cant really fake that stuff via subsidization over long periods without also having all the other pieces of the market in place (all the way from low level workers to capital markets to regulatory environments and even attractive living costs)... otherwise Canada and EU would have grown a larger tech industry by now via their gov programs.
There has to be strong organic production and only then can gov help tip the scales upward, instead of generating it from thin air at the top level. If no one is fleeing China to do it in the US the same way people fled American market conditions to build stuff in Asia it won't happen.
Creating a forced siloed market through tariffs is probably the least efficient and most expensive method to achieving it. But it can plausibly if the domestic market gets used to not buying the nice things the rest of the world has for a decade (similar to what western gov is choosing to do with banning Chinese EVs but applied a thousandfold).
More of a tongue in cheek comment. Gone are the days of "designed" in California "assembled" in China. There's a lot of Chinese designing to get the assembling part done now. If the only way you can goto market in reasonable timeframe is to have PRC iterate your design by rapidly leveraging their supply chains and human capita, then it's not really Made in USA Phone.
How many iterations can they get done in US in 18 months? That's what's going to kill Made in USA, if you can't get the design done on time, not a few $100 extra in PRC vs west sourced BOM, but millions more spent on development over longer time frame because lack of talent. Is the short/medium term solution still to send "homework" to Chinese prototype teams? I suppose economics of PRC speed and few prototypes > 100%+ tariffs.