I will say this gently...You do not know what you are talking about.
https://en.wikipedia.org/wiki/Probe_card
I work in the 'Probe Engineering' dept of a semicon manufacturing company that shall remain nameless. But I will say that I am directly involved in Intel's new 3D CrossPoint NVM product line. By 'directly involved', I mean every wafer has to virtually stop by my desk before it is shipped. I review the functional testing data, via the probe card above, and with proprietary software tools, I say 'nay' or 'yea' for each wafer. Then the shipping dept takes over. That is the overall view.
It literally takes months, around 1/4 to 1/3 of the yr, from wafer start, as in bare silicon wafer, to reach my dept, as in the wafer has fully formed dies ( chips ). Any manufacturing error along the way, assuming the error is detected, will have the wafer, or an entire lot ( batch ) of wafers, scrapped, as in literally toss into the trash barrel.
If the error is human caused, then the next wafer start will have no changes in the manufacturing process. But if the error is not human caused, as in design deficiencies or recipe flaws or equipment inadequacies or many other factors, then the next wafer start must reflect those corrections, so now we are looking at 1/2 of the yr from wafer start to my dept.
If the error is not detected during the manufacturing process and finally detected at the functional testing stage in the Probe dept, then that part of the yr is essentially wasted. Testing may result in zero yield or partial yields of shippable dies per wafer. But no matter which, money is lost.
In all this time, the competition is moving forward while my company struggles to make corrections. They could be making a die size reduction, aka 'shrink', of an existing design, or they could be making a new product line. Using NVM NAND as example, a 64gb die shrink is not a new product line, but a 128gb capacity die is a new product line. Then the whole improvement process starts all over, shrinking that 128gb die.
In semiconductor, if you are consistently one yr behind, you will
NEVER be an innovator. All you will be doing is mass manufacturing of established products. That is what most semicon fabs in mainland China are: contract foundry fabs.
Because of the time involved, China must make acquisition of semicon companies top national priority. Defense cannot rely on imports forever. Whatever semicon products that China can produce at home, as long as those products are technologically behind the West, defense will be behind the West. That is just one example of how semicon affects the nation.
When I used to work for Micron and was working to set up the Micron facility in Shanghai, I was warned of the Chinese propensity for accepting 'good enough' standards, in other words, easily satisfied. This is not racism. This is a carryover from the communist days where the entire culture was suppressed of its creativity. Managers had no to little incentives to produce top quality products of anything since everything of the economy was centrally planned and the bureaucrats were supposedly 'all knowing' and 'all wise'. This mentality permeated every communist country when they existed back in the Cold War days.
We, not just Micron people, had to work extra hard to break Chinese supervisors and managers of that 'good enough' mentality. If the minimum wafer shippable yield is %30, the Chinese were satisfied with that. As long as they meet the wafer shippable weekly quota, everyone, from the production line workers to the top facility executives, were satisfied. We had to tell them the difference between wafer shippable and per wafer yield figures, how they relate to each other, and how each affects the overall financial health of the facility. We had to stress, and there were a lot of hard feelings and terminations, that meeting the per wafer minimum yield is not enough.
We had to show them how and why, that extracting as many functional dies as possible per wafer eventually lower per die cost over the long run, making the facility more cost effective quarter by quarter. Most of the managers never had such financial tools, let alone the necessary business finance education, to be managers, and they were appointed via Party connections so we had to live and work with what we had. Sometimes feelings got so hard that literally physical fights almost broke out. In those instances, we quickly sent our guys home, like the next available flight in the afternoon or even the 'red eye' evening flight, lest national bias and Party power got involved and an American gets jailed.
Here is the kicker -- our Micron people in Taiwan never had such problems. The Taiwanese culture was completely different. If we say the standard is X, they will work hard to meet X and often they will meet expectations within one or two wafer start cycle. If our equipment support supervisor say torque to X, the Taiwanese equipment support tech will get the torque wrench, make sure its calibration is up to date, and use it to meet X. The Chinese tech would have used his elbow's feeling. As a former Air Force guy, I cringe at the idea that a Chinese pilot is flying his jet whose bolts were tightened by someone's elbow's feeling.
Praise China all you want and get all the 'Thanks' you can get from the Chinese in this forum. But I know better.
That is hilarious considering the Soviets and China with their practice of communism was the real hold back of humanity's progress.
