Utilization rate means little by itself. Just because you are operating your factories at high load does not mean you get good yields and viable chips out of it. If you only have 50% yield half of the chips you produce are basically sent to the trash. If the yield is even lower not even that.That's a Samsung design problem with Exynos, not a Samsung foundry problem.
Samsung's 5 nm and 7 nm utilization is 80%-90% though only 50% for other nodes.
...
SMIC utilization rate is also 90%+.
...
That implies demand is not the problem, the problem is capacity.
You are seriously exaggerating the “decades” components of this. The U.S., since it already is a large and wealthy economy, always has capacity in memory (Micron), foundry (Intel, GlobalFoundries), and shipbuilding (Northrop, etc). The issue with the U.S. would be that it would need to scale should an exogenous foreign shock happen. But an exogenous foreign shock would simultaneously increase the profitability of entering the sector, from both a firm and individual perspective so it would at most take a few years. See for example, how quickly supply chains unsnarled themselves and how inflation in the U.S. is back to normal after 16 months despite some very large supply shocks - Covid, Ukraine; interest rate hikes ; and large demand shocks - very large fiscal deficits.This doesn't mean that the loss of Japan and South Korea would not significantly impact the economy of the US. South Korea produces more than 70% of the world's semiconductor memory. Without semiconductor memory, the vast majority of modern electronic devices cannot function. It would take years, if not decades, for the US to come up with a substitute for South Korean semiconductor memory output. Such an endeavor would require tens of thousands of highly-trained professionals. It takes a lot of time to train such a vast and sophisticated workforce. Similarly, the vast majority of cutting-edge semiconductor chips are manufactured in Taiwan. What if the US suddenly lost access to TSMC's facilities? How would investors react if Apple told its shareholders that it would have to halt production for at least 10 years? How would Microsoft's shareholders react if it told them that it could no longer run its cloud business due to a lack of advanced semiconductor chips? How would the US economy react to the resulting changes to the stock market? The TSMC plant in Arizona is not even operational yet. How long would it take for the US to replace TSMC and what opportunity costs would it have to endure during that time? South Korea and Japan also build virtually all of the commercial ships in the world outside of China. Without them, the US would temporarily lose access to the supply of commercial ships. Again, it would take years, if not decades to replace South Korean and Japanese shipbuilding output.
I think it's silly to assume that the US would be completely fine on its own.
Let's agree to disagree. If you are actually from Hillsboro and work for Intel, I would be surprised that you think it would take anything less than decades for the US to completely replace its semiconductor imports, especially if it is cut off from the rest of the world.You are seriously exaggerating the “decades” components of this. The U.S., since it already is a large and wealthy economy, always has capacity in memory (Micron), foundry (Intel, GlobalFoundries), and shipbuilding (Northrop, etc). The issue with the U.S. would be that it would need to scale should an exogenous foreign shock happen. But an exogenous foreign shock would simultaneously increase the profitability of entering the sector, from both a firm and individual perspective so it would at most take a few years. See for example, how quickly supply chains unsnarled themselves and how inflation in the U.S. is back to normal after 16 months despite some very large supply shocks - Covid, Ukraine; interest rate hikes ; and large demand shocks - very large fiscal deficits.
what’s more: Japanese and South Korean firms export to the United States in order to make money themselves; without final American demand, those firms lack the scale to compete effectively
Capacity is a problem for sure, so if TSMC USA and Intel are adding sub 10nm capacity, that would only be assuredly fully utilized (in the Taiwan go boom scenario).That's a Samsung design problem with Exynos, not a Samsung foundry problem.
Samsung's 5 nm and 7 nm utilization is 80%-90% though only 50% for other nodes.
SMIC utilization rate is also 90%+.
That implies demand is not the problem, the problem is capacity.
I think you are underestimating importance of Japan and Korea. start with display technology to vehicle tires to batteries to vehicle climate control. and i highly doubt any one can make as reliable car transmission/ engines as Japanese. US cannot even replace Indian software engineers when it comes to scale let alone Japan/Korea in manufacturing and it is getting worse in things that are important for Mideast like recent air defense deals.You are seriously exaggerating the “decades” components of this. The U.S., since it already is a large and wealthy economy, always has capacity in memory (Micron), foundry (Intel, GlobalFoundries), and shipbuilding (Northrop, etc). The issue with the U.S. would be that it would need to scale should an exogenous foreign shock happen. But an exogenous foreign shock would simultaneously increase the profitability of entering the sector, from both a firm and individual perspective so it would at most take a few years. See for example, how quickly supply chains unsnarled themselves and how inflation in the U.S. is back to normal after 16 months despite some very large supply shocks - Covid, Ukraine; interest rate hikes ; and large demand shocks - very large fiscal deficits.
Japan and Korea are using American globalized labor and raw materials for there own benefit and have a sway over US.what’s more: Japanese and South Korean firms export to the United States in order to make money themselves; without final American demand, those firms lack the scale to compete effectively