Chinese semiconductor industry

FriedRiceNSpice said:

Semiconductor manufacturing has some similarities with aircraft engine manufacturing. In both cases, it can be extremely difficult and time consuming to go from understanding the science and being able to build a prototype in the lab to having a viable manufacturing process for mass production. Both cases involve many cycles of trial and error and refinement. A good example from aircraft engine manufacturing would be the single-crystal engine blades, where even after successful prototypes can be made in labs, it still takes many years to develop a mass manufacturing process that can produce acceptable yields. Semiconductor manufacturers face the exact same issue, which is how to refine the manufacturing process so that the yields are commercially viable. It is no coincidence that while China has managed to catch up to the rest of the world in so many different fields, aero engines and semiconductors are the two main fields that China is still lagging behind in.

Click to expand...

Your analogy can literally apply to any manufacturing process. If you took a operations management course, you would know throughput matters for making anything. Whether it's gas turbines or semiconductors or shoes.

These two industries have very little in common. One is based around advanced metallurgy. Another is dependent on perfecting the physics behind light and lasers.

China has much more talent in the latter than metallurgy. And is in fact leading in the area of high powered lasers which is the backbone of lithography machines.

So you are wrong. Simple analogies without in depth analysis tells us nothing.

Also consider this, gas turbine blades are exposed to a lot more different variables than making semiconductors. Heat, pressure, rotational forces, metallurgy, casting process temperatures, etc.

Lithography is done in a vacuum. There aren't as many variables which affect product performance or yield. So it takes less time to perfect.

free_6ix9ine said:

Your analogy can literally apply to any manufacturing process. If you took a operations management course, you would know throughput matters for making anything. Whether it's gas turbines or semiconductors or shoes.

These two industries have very little in common. One is based around advanced metallurgy. Another is dependent on perfecting the physics behind light and lasers.

China has much more talent in the latter than metallurgy. And is in fact leading in the area of high powered lasers which is the backbone of lithography machines.

So you are wrong. Simple analogies without in depth analysis tells us nothing.

Click to expand...

Saying any manufacturing process takes time to perfect and has yield targets is a truism - it does not take decades to develop a manufacturing process with acceptable yields for manufacturing shoes. The manufacturing processes for semiconductors and aircraft engines are notoriously difficult to develop. The fact that they are from completely dissimilar industries / scientific disciplines does not negate that fact. I admit I am personally don't have the in-depth knowledge to give a deeper analysis, perhaps someone with more relevant experience or knowledge on both fields would care to elaborate on the specific challenges that make each manufacturing process so difficult to perfect.

FriedRiceNSpice said:

Semiconductor manufacturing has some similarities with aircraft engine manufacturing. In both cases, it can be extremely difficult and time consuming to go from understanding the science and being able to build a prototype in the lab to having a viable manufacturing process for mass production. Both cases involve many cycles of trial and error and refinement.

Click to expand...

First off, everything worth pursuing requires time and trial-error. This is not unique to semiconductors, otherwise, Taiwan/Korea/Netherlands would be making their jet engines too. Which leads me to my second point....

FriedRiceNSpice said:

It is no coincidence that while China has managed to catch up to the rest of the world in so many different fields, aero engines and semiconductors are the two main fields that China is still lagging behind in.

Click to expand...

Second, every country needs to specialize and cannot be the master-of-all-things. Taiwan/Korea/Netherlands buy their jet engines from elsewhere. US/UK/FR/RUS/CHN buy their leading nodes from TW/KOR.

There is only a limited number of engineers, that's why Netherlands specializes in equipment, Taiwan specializes in pure play foundry, USA specializes in design. It's the law of limited resources.

Can you name a single country that has leading jet engine tech and leading nodes? (Don't say US, because Intel abdicated the throne) Saying China lacks leading tech in jet engines & leading nodes, well, EVERY COUNTRY lacks that too. Saying the most obvious isn't really insightful.

I believe China can eventually match parity because Moore's law is slowing down, which gives China more time to catch-up.

FriedRiceNSpice said:

Saying any manufacturing process takes time to perfect and has yield targets is a truism - it does not take decades to develop a manufacturing process with acceptable yields for manufacturing shoes. The manufacturing processes for semiconductors and aircraft engines are notoriously difficult to develop. The fact that they are from completely dissimilar industries / scientific disciplines does not negate that fact. I admit I am personally don't have the in-depth knowledge to give a deeper analysis, perhaps someone with more relevant experience or knowledge on both fields would care to elaborate on the specific challenges that make each manufacturing process so difficult to perfect.

Click to expand...

Gas turbines is primarily figuring out the right mixture of nickel alloys to overcome high temperatures and cleverily designing your engine to funnel air to cool the blades so they don't melt. There are a thousand variables that affect the end performance of your engine. That's why you need to collect a ton of data from trial and error experiments and analyze it to determine what works and what doesnt. That's why it takes so long to develop an engine.

Lithography which is the weakest link for China is playing with light to produce the highest resolution of images with minimal defects. Of course you need trial and error, but the amount of variables are less, since chips operate in more or less ambient environment.

Most importantly it's much easier to run lithography experiments than run gas turbine experiments. Imagine the production process for a new alloy, you have to develop the alloy, cast thousand of blades, assemble it properly in the engine, etc. It might take 1yr to run a gas turbine experiment, but a lithography experiment can be done in a week. Because the process is less involved.

Never mind the fact that machine learning has vastly improved the ability to analyze data.

free_6ix9ine said:

Your analogy can literally apply to any manufacturing process. If you took a operations management course, you would know throughput matters for making anything. Whether it's gas turbines or semiconductors or shoes.

These two industries have very little in common. One is based around advanced metallurgy. Another is dependent on perfecting the physics behind light and lasers.

China has much more talent in the latter than metallurgy. And is in fact leading in the area of high powered lasers which is the backbone of lithography machines.

So you are wrong. Simple analogies without in depth analysis tells us nothing.

Click to expand...

The two do have a lot in common. A more apt comparison would be how aircraft engines conceptually use thermodynamics and photolithography quantum physics, but their manufacture requires other technologies that have nothing to do with the initial science.

To properly exploit thermodynamics to create a high-performing engine, you need advanced alloys to create the blades that operate under high-temperature environments. Similarly, to properly exploit quantum physics to create a high-performing photolithography machine, you need high-precision machining to create the optics that focus the light onto the wafer, and many more.

That's not all. Once you have a tech demonstrator, the true challenge begins: how to do it again, tens or even hundreds of times over, quickly, reliably, economically. While I have no doubt that China has the capability to succeed, saying that photolithography is somehow easier than jet engines because of perceived complexity is just wishful thinking and, frankly, insulting to the researchers hard at work in SMEE.

That is not correct take single crystal blade. You manufactured it in tested it in the lab condition It work perfect Then you build the actual engine test it in bench condition It work perfect. Then tested it actual flying condition or nowadays in simulated weather and altitude condition. It does not worked.

So you have to start all over again. That is why it take so long what it work in lab condition might not work in real time condition Because there are so many variable The western way of doing it is to built engine first since the engine and the fighter jet manufacturer are not the same company . Many years before they even had the fighter on paper. The Chinese built the engine and the fighter at the same time That is why it give the impression that they have lagged or incapable of building turbofan .But now that they separate the 2 It should give more progress.

Chinese academy science has prepared the leg work for a long time So they already has the basic science figure it out Now it is to get all the components and the financial back up and most important the FAB cooperation in figuring the right integration

Hendrik_2000 said:

That is not correct take single crystal blade. You manufactured it in tested it in the lab condition It work perfect Then you build the actual engine test it in bench condition It work perfect. Then tested it actual flying condition or nowadays in simulated weather and altitude condition. It does not worked.

So you have to start all over again. That is why it take so long what it work in lab condition might not work in real time condition Because there are so many variable The western way of doing it is to built engine first since the engine and the fighter jet manufacturer are not the same company . Many years before they even had the fighter on paper. The Chinese built the engine and the fighter at the same time That is why it give the impression that they have lagged or incapable of building turbofan .But now that they separate the 2 It should give more progress.

Chinese academy science has prepared the leg work for a long time So they already has the basic science figure it out Now it is to get all the components and the financial back up and most important the FAB cooperation in figuring the right integration

Click to expand...

I agree. The environment a lithography machine operates in a fab is almost the same as in a lab. It's not subjected to environmental changes such as air, fuel quality, temperatures, Contaniments. once you perfect a lithography machine in the lab, it's closer to completion than if you perfect a turbine blade in a lab environment or testing environment.

Phead128 said:

Can you name a single country that has leading jet engine tech and leading nodes? (Don't say US, because Intel abdicated the throne) Saying China lacks leading tech in jet engines & leading nodes, well, EVERY COUNTRY lacks that too. Saying the most obvious isn't really insightful.

Click to expand...

Yes, very few countries can do semiconductors and jet engines. That is the point - it is extremely difficult and requires very large investments of time and resources. China has the resources to master these fields, but it is still going to take time. With engines, since China started earlier, it is already close to reaping the fruits of its investment, but it did take a very long time. You are right with Moore's law, it means at least China does not have to chase a moving target, which does make it easier to catch-up.

Hendrik_2000 said:

That is why it give the impression that they have lagged or incapable of building turbofan .But now that they separate the 2 It should give more progress.

Click to expand...

It illustrates the relative greater time and investment that turbofans take relative to the other technologies/processes that China had to catch-up in with regards to fighter development. On the other hand, China caught up in terms of electronics / radars / avionics in a very short period of time.

Hendrik_2000 said:

That is not correct take single crystal blade. You manufactured it in tested it in the lab condition It work perfect Then you build the actual engine test it in bench condition It work perfect. Then tested it actual flying condition or nowadays in simulated weather and altitude condition. It does not worked.

Click to expand...

From what I have heard, there was also a bottleneck with reaching acceptable yields with the manufacturing process. The proportion of single-crystal blades being produced that passed quality checks was too low to be viable for mass production.

FriedRiceNSpice said:

It illustrates the relative greater time and investment that turbofans take relative to the other technologies/processes that China had to catch-up in with regards to fighter development. On the other hand, China caught up in terms of electronics / radars / avionics in a very short period of time.

From what I have heard, there was also a bottleneck with reaching acceptable yields with the manufacturing process. The proportion of single-crystal blades being produced that passed quality checks was too low to be viable for mass production.

Click to expand...

No metallurgy is more an art than science basically trial and error The west has advantage since they started it a long time ago and accumulated experience and skill over time America just take over this experience and skill. China has to figure it out by herself

Laser and plasma technology is more of science than experiment there is formula there huge knowledge that you can apply like computer simulation computer calculation, physical theory etc. And as free guy said very limited environment condition It all play to China strength in Physics and Math

China is now mass producing WS-10 engines. It is just a matter of time before WS-15 are mass produced. Most likely they will achieve this within a shorter period of time compared with time needed for the WS-10.