Indian Electronics Manufacturing Developments : News, Updates and Discussions

Also, what makes you think India is going to respect international copyright laws if the country is sanctioned? Although a total sanction is unlikely, a Huawei type ban is more likely and feasible.
To avoid a Huawei type of ban (which you also agree is more likely), it is absolutely necessary to develop our own IP-core (synthesizable HDL implementations, microarchitecture etc) pool. Our own EDA software toolchain. We can even support open source projects which are implementing this. There are many such projects running via universities etc. It is paramount to have IP-core and tool chains free from US government interefence.

Just think if Linux and Apache webserver was not under open licenses. US could have a ability to block India or China out of OS and webserver software ecosystem by restricting likes of IBM, erstwhile Sun Microsystem, Microsoft etc from selling those to Indian companies.

Same is possible with IP-cores of processors, SoCs etc. One area which needs absolute attention is EDA tools like synthsis tools, PCB CAD tools, place and route tools, power analysis tools. There are not good GNU replacements for these and all of these are softwares. US is able to pressure China using these tools. Cadence stopped delivering updates to Huawei's synthesis software.

This is what needs to be defeated. It does not require manufacturing but design and software development.
enuf of BS, there is no point in discussing with you. Iran, NK are good examples of what US can do.
LOL! If NK and Iran level sanctions are hitting India, worry more about oil and banking and less about chips. Anyways cya and thanks for rage quitting :)
 
Errr... That thing actually exists in India. Believe it or not, but India had been at forefront of design competency for quite sometime. Quite a few of the chipset ASIC in intel board were actually designed and taped out of Sarjapur Ring Road campus of Intel. There was actually a completely designed-in-India server processor called "Project Whitefield" back in 2000s before it got cancelled due to political reasons. Ask @Bali78, he might know more.

IIT-Delhi runs a program called VDTT (https://vdtt.iitd.ac.in/). And there is similar one at IIT-B. These are more than a decade old. This part is not as big of a problem. Problem is that India never matured from a design service to a product powerhouse AFAIK.

I already know that. So it's not a problem from the design end. It's a problem from the production side.

Once more, design is not an issue. Not at all. Fabrication in India at scale is just not worth it.

Depends on who you're catering to. For example, let's say new smartphones are made that largely use Indian IP, and is cheap enough to be sold to a whole lot of third world countries, not just India. The Indian IP for the hardware ensures significant amounts of security for Indians, while being cheap enough to be used by the third world. Even if it's not exported, there will still be room for India because the market is over a billion strong.

And as I've already stated, the fab plants are needed for everything, not just smartphones and tablets, so we are talking about electrical appliances, vehicles, communication systems, military etc. We are talking about many billions of chips a year. I have no clue what you mean by scale when the plants are expected to deliver semiconductor hardware to a retail industry that's gonna be worth a few trillions of dollars by 2030.
 
You cann't replace Oil for quite sometime. In short to medium time, your pharma will be toast. You will be locked out of global banking system completely. Medicine prices will be sky high and you won't even be able to receive money from outside. That level of sanction means a absolute and total failure of our Diplomacy. At that point, semiconductor for phones, computer will be a far cry. I will worry if my bike or car will have petrol OR the trucks will be running to ship me food. Priorities..

None of that's gonna happen today or even a few years from now. The potential for sanctions is in the future, after 2030. As I've already pointed out before, India won't even be an upper middle income country until after 2035 at 2020 prices.
 
To avoid a Huawei type of ban (which you also agree is more likely), it is absolutely necessary to develop our own IP-core (synthesizable HDL implementations, microarchitecture etc) pool. Our own EDA software toolchain. We can even support open source projects which are implementing this. There are many such projects running via universities etc. It is paramount to have IP-core and tool chains free from US government interefence.

Just think if Linux and Apache webserver was not under open licenses. US could have a ability to block India or China out of OS and webserver software ecosystem by restricting likes of IBM, erstwhile Sun Microsystem, Microsoft etc from selling those to Indian companies.

Same is possible with IP-cores of processors, SoCs etc. One area which needs absolute attention is EDA tools like synthsis tools, PCB CAD tools, place and route tools, power analysis tools. There are not good GNU replacements for these and all of these are softwares. US is able to pressure China using these tools. Cadence stopped delivering updates to Huawei's synthesis software.

This is what needs to be defeated. It does not require manufacturing but design and software development.

Yep. If we are hit by this type of sanctions, then we can only circumvent it using our own IP, regardless of where production happens. So this affects only the company. But the country is in a different situation because of forex. Wherever the new products are made, we have to import it using dollars, dollars that we are unlikely to have at the scale necessary.

Think about it. Our forex reserves today are around $600B, it's not going to climb to $3T within the next decade for obvious reasons, so it's at best gonna be $1-1.5T in total. But our electronics industry alone is expected to require forex well above $1T every single year. Do you see what I'm getting at? Our forex must be used only for natural resources. We need oil, gold and rare-earth metals imported. We can't waste forex on electronics when it can be made at home, even if it's more expensive.

But at the same time, we need to prepare for conditions where full sanctions are possible. Hence the need to indigenise for the sake of security as well as controlling forex outgo.
 
Errr... That thing actually exists in India. Believe it or not, but India had been at forefront of design competency for quite sometime. Quite a few of the chipset ASIC in intel board were actually designed and taped out of Sarjapur Ring Road campus of Intel. There was actually a completely designed-in-India server processor called "Project Whitefield" back in 2000s before it got cancelled due to political reasons. Ask @Bali78, he might know more.

IIT-Delhi runs a program called VDTT (https://vdtt.iitd.ac.in/). And there is similar one at IIT-B. These are more than a decade old. This part is not as big of a problem. Problem is that India never matured from a design service to a product powerhouse AFAIK.


Once more, design is not an issue. Not at all. Fabrication in India at scale is just not worth it.
Yes. Whitefield was completely designed by intel India and later cancelled. TI India designed many mobile SoCs during 2000 -2010 period.

As far as no of designers are concerned, right now there are way more Indian chip designers than Americans. In Bay Area, all chip design groups are dominated by Asians. I haven’t interviewed a single white guy in past many years!!
 
  • Like
Reactions: Proud_Indian
I already know that. So it's not a problem from the design end. It's a problem from the production side.
Its weird actually. We ie Indians tend to be a bit slow in putting our products in market. Most of the products that the best engineers among us build are usually of the type "Built by Indians in India BUT not Made in India" kind of products. This is true across the board. There are way too many quality Indian software engineers, even many are supporting many opensource projects BUT you wont find too many Indian companies actually selling "Made in India" softwares. They usually sell "Get it Built in India" kind of services : like outsourcing or offshore development.

It shows up in IP-core development equally. There are many companies or subsidairies that will take an outsourced project and will implement is pretty decently, but on one will dare implement their own ip-cores to sell.

I guess it might comes from a psychological block of extreme and over avoidance of risk. I call it "Anti-Enterpenureship". Even the promise that one will be paid at the end of project is more than enough for us to work but we won't work if there is a chance that product will run a risk of not being bought in market.

@Bali78 are you aware of any Indian companies selling IPs (cores/toolchain) like Broadcom or other fabless design houses do?
 
Depends on who you're catering to. For example, let's say new smartphones are made that largely use Indian IP, and is cheap enough to be sold to a whole lot of third world countries, not just India. The Indian IP for the hardware ensures significant amounts of security for Indians, while being cheap enough to be used by the third world. Even if it's not exported, there will still be room for India because the market is over a billion strong.

And as I've already stated, the fab plants are needed for everything, not just smartphones and tablets, so we are talking about electrical appliances, vehicles, communication systems, military etc. We are talking about many billions of chips a year. I have no clue what you mean by scale when the plants are expected to deliver semiconductor hardware to a retail industry that's gonna be worth a few trillions of dollars by 2030.
IP can be made in India, smart phone can be made in India but chips will likely be fabricated in Taiwan...

To be honest if that happens, I will call it an absolute success. If an Indian company builds a SoC around Shakti Processor with good number of Indian IPs and some licensed IPs from South Korea or Taiwan; it will be a great success. If we have a SoC platform like Hauwei's Kirin with RISC-V (rather Shakti implementation) and a tablet or smartphone around it, then its a massive breakthrough. Does not matter if the fabrication is done in Taiwan.
Think about it. Our forex reserves today are around $600B, it's not going to climb to $3T within the next decade for obvious reasons, so it's at best gonna be $1-1.5T in total. But our electronics industry alone is expected to require forex well above $1T every single year. Do you see what I'm getting at? Our forex must be used only for natural resources. We need oil, gold and rare-earth metals imported. We can't waste forex on electronics when it can be made at home, even if it's more expensive.

Think it in this way : China imports 100s of billion dollars of chips but that does not put strain on their forex. Why? Because they build finished goods using them and re-export them. Its value addition which results in net positive impact on forex EVEN if they import a lot of semiconductor. It helps them offer latest and greatest chips in their products and at a very low price.
We do the same with petro.

For Strategic projects, we have CSL with 180 nm nodes for limited production. That IMHO is a good place. I will like it to be scale up to 90 nm and later 65 nm at the right time. That will provide us with enough capability to ensure chips going into Astra/ISRO are not compromised. A lot can also be done by right application of FPGAs and CPLDs too.
 
As far as no of designers are concerned, right now there are way more Indian chip designers than Americans. In Bay Area, all chip design groups are dominated by Asians. I haven’t interviewed a single white guy in past many years!!
EE is declining in the west among native westerners. Mostly Chinese and Indians take up EE in the west.
 
IP can be made in India, smart phone can be made in India but chips will likely be fabricated in Taiwan...

To be honest if that happens, I will call it an absolute success. If an Indian company builds a SoC around Shakti Processor with good number of Indian IPs and some licensed IPs from South Korea or Taiwan; it will be a great success. If we have a SoC platform like Hauwei's Kirin with RISC-V (rather Shakti implementation) and a tablet or smartphone around it, then its a massive breakthrough. Does not matter if the fabrication is done in Taiwan.

As I've pointed it out before, we can't afford imports. If imports are done, then the govt will be forced to add anywhere between 10 and 100% in tariffs in order to force people from not using it.

Think it in this way : China imports 100s of billion dollars of chips but that does not put strain on their forex. Why? Because they build finished goods using them and re-export them. Its value addition which results in net positive impact on forex EVEN if they import a lot of semiconductor. It helps them offer latest and greatest chips in their products and at a very low price.
We do the same with petro.

Been telling you that you can't compare China and India. China is a significantly more mature market, their retail space is worth $5.6T, ours is $800B. Their forex reserves are over $3T. At $10000 China is at the peak of an upper middle income country ($4000-$12500). At $1800-2000 we are not even at the peak of a lower middle income country ($1000-$4000). The quality of their economy is something else entirely.

What we do with petro can't be done with electronics because the scale is completely different and the competition is also different. When it comes to oil, there is a market for everybody, but when it comes to finished goods, the market is only for quality products. So when we speak of scale, our market today is very small, which is why we only consume 4 million bpd. Refined petroleum products depends on external markets and the scale of our exports is very small in comparison to our future needs. It works out today only because our market is small today, so there is some kind of balance between our crude oil imports and refined oil exports. But as the internal market grows bigger, the difference in imports of crude and exports of refined will keep growing. And when it comes to electronics, the difference is going to be mind numbingly large simply because other countries have superior IP and better product support, all supported by superior markets, so we won't be able to compete.

China did it with smartphones because their internal market matured to such an extent relatively quickly. And they made products specifically for the third world and it resonated with the third world. We do not yet have the foundation for that, and by the time we get a foundation, our electronics imports will completely stall due to affordability.

I still don't think you have understood. The govt will be forced to raise taxes or ban imports if we can't afford electronics. TSMC may make it cheaper, but if we can't afford it in dollars, the govt will introduce capital controls anyway, making mainstream electronics out of reach of the masses. This has nothing to do with profit and loss, it has everything to do with strategic compulsions and current account. As already proven, the govt will run a company at a tremendous loss, like BSNL's losses going into many billions, as long as there is a need for the industry. Literally nothing to do with making profits. Our forex today is almost entirely dependent on foreign support.

For Strategic projects, we have CSL with 180 nm nodes for limited production. That IMHO is a good place. I will like it to be scale up to 90 nm and later 65 nm at the right time. That will provide us with enough capability to ensure chips going into Astra/ISRO are not compromised. A lot can also be done by right application of FPGAs and CPLDs too.

Strategic projects have fab plants already, they do not need anymore addition. There are 4. 2 are silicon (Chandigarh and Bangalore), 1 GaAs (Hyderabad) and 1 GaN (Bangalore).

GoI's fab plant proposal has three options. First is a greenfield project, second is an expansion of existing strategic plants for commercial use, third is the purchase of a foreign plant. So GoI has sent out proposals to the industry and expect a reply back by the end of Jan.
 
Been telling you that you can't compare China and India. China is a significantly more mature market, their retail space is worth $5.6T, ours is $800B. Their forex reserves are over $3T. At $10000 China is at the peak of an upper middle income country ($4000-$12500). At $1800-2000 we are not even at the peak of a lower middle income country ($1000-$4000). The quality of their economy is something else entirely.
I don't see the point of bring China's entire domestic retail market. China's overall ICs import is $150 billion. If we were importing 50% of that in ICs and building equipments locally, we wont need $3 Trillion of forex reserve cushion, our current forex reserve should be enough. Bonus if you use that forex cushion to depreciate Indian currency to make exports of Indian produced good even more attractive.

I mean, for example, my son uses a C$30 toy that is made in China. It has a single color display and a microcontroller running it and producing all sorts of sounds. I checked, the microcontroller costs less than 50 cents in more than 1000+ unit order (from no name Taiwanese company Padauk) and the display about 2-3 dollars (US $1.95 10% OFF|LCD Display Module Monitor White Backlight Adapter PCB 84*48 84x48 5110 Screen for Arduino Controller 3.3V Dot Matrix Digital|screen for arduino|lcd modulemodule display - AliExpress). Total cost of electronics is less than 5 dollars in bulk. Assuming a 100% markup, it will be a shame if in India we cann't make a plastic molded casing and pcb for 5-6$s a piece.

If we cann't, we don't need to a fab etc. We need competitive BASIC manufacturing capability. You know molding cases and fitting electronics and making basic single layer PCBs but on cheap and on scale. Not the HAL style manufacturing.

1608793842846.png


What we do with petro can't be done with electronics because the scale is completely different and the competition is also different. When it comes to oil, there is a market for everybody, but when it comes to finished goods, the market is only for quality products.
Well a trip around walmart here suggests there is space for everyone. Dollar shop, anyone?
The shame is that we cann't even produce the shit that is being sold in those places on cheap. For example, the nick-nack above. It sells a lot and its not a lot of engineering. Just plastic molding, fabricating PCB and putting them together. Do you know the real trouble, why it never happens in India?

Here it is. The biggest plastic molding company in India, likes of Neel-Kamal are here :
  • North – Samba (Jammu & Kashmir) and Greater Noida (Uttar Pradesh)
  • South – Pondicherry (Union Territory)
  • East – Barjora (West Bengal)
  • West – Sinnar, Nashik (Maharashtra) and Silvassa (Union Territory of Dadra & Nagar Haveli) (2 plants).
The biggest PCB makers? New Delhi, Mumbai, Pune.

You go to China and ShenDingDong city, they are all located in ONE campus. You go to a plastic case modler for your case production and testing, you go to the PCB fabricator next door, they make your PCBs. You can setup a reliable manufacturing system in a week by signing contracts. In few months you will have your goods done ready for shipment.

If you do the same in India, nothing will happen. You need to fabricate PCBs in Delhi, get plastic molded in Kolkatta and manage a Farmer's protest which can delay your orders by 2 months.

No amount of semiconductor fabrication will fix above. All the semi-conductors can go "F" themselves if you don't have these basic streamlined manufacturing parks or campuses. Lets not kid ourselves. The Modi government is doing one more blunder if it is planning to develop any kind of large scale chip fabrication capability in India before basic chops to manufacture simple goods.

Somehow, the obvious things are never done in India. All the weird things are done first.
 
I still don't think you have understood. The govt will be forced to raise taxes or ban imports if we can't afford electronics. TSMC may make it cheaper, but if we can't afford it in dollars, the govt will introduce capital controls anyway, making mainstream electronics out of reach of the masses. This has nothing to do with profit and loss, it has everything to do with strategic compulsions and current account. As already proven, the govt will run a company at a tremendous loss, like BSNL's losses going into many billions, as long as there is a need for the industry. Literally nothing to do with making profits. Our forex today is almost entirely dependent on foreign support.
"Mainstream Electronics". LOL.

IF it comes to that, then why do us Indian the "Mainstream CONSUMER Electronics". I mean I recently "upgraded" to iPhone SE 1 from iPhone 5S. Thats the LAST iPhone available in the form factor I love. I hate the massive bricks they sell as phones these days. It was a pre-owned phone and the store cleaned it up really well and gave me a warrenty of 1 year on it. Total price? 100 canadian dollars.

IF Indian currency keeps on declining, this is what we should do then. There will ALWAYS be an excess of such "Mainstream CONSUMER Electronics". Heck, you can bring just their innards and manufacture quality tropicalized cases around those in India. Locally manufacture LiPo batteries in India. Done in bulk you can have quality electronics for REAL cheap with local manufacturing.

The need for new electronics is actually NOT as much as people think it is.


That being said, if we fix our exports then we do not need to do the above. The problem in India is NOT as much as imports. Its falling exports that will put more load on our forex. Our exports fall because our manufacturing is unorganized mess.
 
I don't see the point of bring China's entire domestic retail market. China's overall ICs import is $150 billion. If we were importing 50% of that in ICs and building equipments locally, we wont need $3 Trillion of forex reserve cushion, our current forex reserve should be enough. Bonus if you use that forex cushion to depreciate Indian currency to make exports of Indian produced good even more attractive.

The Chinese have tons of fab plants already, which is why their overall IC imports is only $150B.

And their future looks good.
China has the most fab projects in the world, with 30 new facilities or lines in construction or on the drawing board, according to data from SEMI’s World Fab Forecast Report. Of those, 13 fabs are targeted for the foundry market, according to SEMI. The remaining facilities are geared toward LEDs, memory and other technologies.

Still, China continues to build up foundry capacity. “Out of the 30 fabs and lines starting construction in 2018 and beyond, 13 are for foundry,” said Christian Dieseldorff, an analyst at SEMI. “As of the end of 2018, our data showed an installed foundry capacity of 1.3 million wpm in 200mm equivalents in China. Worldwide is about 7 million wpm. This includes IDM foundry capacity.”


Well a trip around walmart here suggests there is space for everyone. Dollar shop, anyone?
The shame is that we cann't even produce the shit that is being sold in those places on cheap. For example, the nick-nack above. It sells a lot and its not a lot of engineering. Just plastic molding, fabricating PCB and putting them together. Do you know the real trouble, why it never happens in India?

Here it is. The biggest plastic molding company in India, likes of Neel-Kamal are here :
  • North – Samba (Jammu & Kashmir) and Greater Noida (Uttar Pradesh)
  • South – Pondicherry (Union Territory)
  • East – Barjora (West Bengal)
  • West – Sinnar, Nashik (Maharashtra) and Silvassa (Union Territory of Dadra & Nagar Haveli) (2 plants).
The biggest PCB makers? New Delhi, Mumbai, Pune.

You go to China and ShenDingDong city, they are all located in ONE campus. You go to a plastic case modler for your case production and testing, you go to the PCB fabricator next door, they make your PCBs. You can setup a reliable manufacturing system in a week by signing contracts. In few months you will have your goods done ready for shipment.

If you do the same in India, nothing will happen. You need to fabricate PCBs in Delhi, get plastic molded in Kolkatta and manage a Farmer's protest which can delay your orders by 2 months.

No amount of semiconductor fabrication will fix above. All the semi-conductors can go "F" themselves if you don't have these basic streamlined manufacturing parks or campuses. Lets not kid ourselves. The Modi government is doing one more blunder if it is planning to develop any kind of large scale chip fabrication capability in India before basic chops to manufacture simple goods.

Somehow, the obvious things are never done in India. All the weird things are done first.

That's actually the plan. Hyderabad had created an SEZ called FAB City, which is exclusively for the semiconductor business. More such SEZs are planned under the GoI's resurrected fab plans.


"Mainstream Electronics". LOL.

IF it comes to that, then why do us Indian the "Mainstream CONSUMER Electronics". I mean I recently "upgraded" to iPhone SE 1 from iPhone 5S. Thats the LAST iPhone available in the form factor I love. I hate the massive bricks they sell as phones these days. It was a pre-owned phone and the store cleaned it up really well and gave me a warrenty of 1 year on it. Total price? 100 canadian dollars.

IF Indian currency keeps on declining, this is what we should do then. There will ALWAYS be an excess of such "Mainstream CONSUMER Electronics". Heck, you can bring just their innards and manufacture quality tropicalized cases around those in India. Locally manufacture LiPo batteries in India. Done in bulk you can have quality electronics for REAL cheap with local manufacturing.

The need for new electronics is actually NOT as much as people think it is.


That being said, if we fix our exports then we do not need to do the above. The problem in India is NOT as much as imports. Its falling exports that will put more load on our forex. Our exports fall because our manufacturing is unorganized mess.

Easier said than done. The primary goal is self-sufficiency. For exports, we need to make a competitive industry first. Have you considered we are incapable of playing the export game? There are plenty of countries where manufacturing is not an unorganised mess.

Our main goal is to sustain this.

Our people are going to remain far too poor for a long time which will prevent the reforms needed to make us competitive outside the country. Look at the farmer agitation happening now. Even 10 years down the line, it's going to be the same. The massive disparity in the size of our domestic market versus our electronics exports shows we are a long way away from doing what you said.

The business of selling refurbished products is big is India, but is absolute shit. Even then that $100 phone still has to be imported. An import we won't be able to afford due to the sheer scale of the market.

I still don't think you appreciate how big the Indian market really is and why we can't afford the imports of electronics.
 
That's actually the plan. Hyderabad had created an SEZ called FAB City, which is exclusively for the semiconductor business. More such SEZs are planned under the GoI's resurrected fab plans.
And thats exactly why all these will fail. Badly.
India wants to be another Korea or Taiwan or Japan in semiconductor business. It will get burnt. It has happened before in 2000s, it will happen again.
They don't need these stupid FAB SEZs. They need finished goods SEZs.
Last time when they established a Pharma SEZ in Pune, it paid off well. Margins were decent.
Now they want to go for super low margin goods, devil knows why. It will be another STM fab debacle of 2000s.

China has the most fab projects in the world, with 30 new facilities or lines in construction or on the drawing board, according to data from SEMI’s World Fab Forecast Report. Of those, 13 fabs are targeted for the foundry market, according to SEMI. The remaining facilities are geared toward LEDs, memory and other technologies.

Still, China continues to build up foundry capacity. “Out of the 30 fabs and lines starting construction in 2018 and beyond, 13 are for foundry,” said Christian Dieseldorff, an analyst at SEMI. “As of the end of 2018, our data showed an installed foundry capacity of 1.3 million wpm in 200mm equivalents in China. Worldwide is about 7 million wpm. This includes IDM foundry capacity.”
China is doing that AFTER undergoing basic manufacturing revolution in 90s. When China was about the size of Indian economy ie in 2002 or 03, they were busy enticing less capital intensive manufacturing to China. India is going to burn its fingers with a very high capital intensive business.
Easier said than done. The primary goal is self-sufficiency. For exports, we need to make a competitive industry first. Have you considered we are incapable of playing the export game? There are plenty of countries where manufacturing is not an unorganised mess.
Simple reason, cost of manufacturing. India has a massive potential low cost labour pool to build factories producing massive amounts of goods :electrical electronic or anything. The only other country which can do so is China.
Name one country which has that potential and a very oragnized manufacturing sector. There are none.
India needs to get its organized manufacturing in place first while maximizing the utility of capital. Once they have industries set up that can take the space getting vacated by China now, we can think of all these competing part. India's competition is with Vietnam and Bangladesh. Beat them at this game of manufacturing and then think of Fabs etc etc etc.
 
Last edited:
I still don't think you appreciate how big the Indian market really is and why we can't afford the imports of electronics.
What you fail to appriciate is that people do not buy semiconductors. They buy electronic goods. It is possible to have a massive electronic industry WITHOUT wasting money in setting up these white elephants called fabs.
Manufacture goods that sell. Manufacture PCBs, mechanical parts and finished goods. Once you have 100s of billions to burn, then establish your own top of the line fab industry. China did the same.
 
Simple reason, cost of manufacturing. India has a massive potential low cost labour pool to build factories producing massive amounts of goods :electrical electronic or anything. The only other country which can do so is China.
Name one country which has that potential and a very oragnized manufacturing sector. There are none.
India needs to get its organized manufacturing in place first while maximizing the utility of capital. Once they have industries set up that can take the space getting vacated by China now, we can think of all these competing part. India's competition is with Vietnam and Bangladesh. Beat them at this game of manufacturing and then think of Fabs etc etc etc.

By the time we "beat" Vietnam, BD, even China, we are gonna broke importing $200-300B worth of IC chips in 2030-35. We can't go to the IMF for a bail-out package just for ICs.

Thankfully, India is not going to follow what you said. GoI is pretty sure about the fab plants now. If we don't get access to the international market for the fab plant, then we cater to the domestic market only, even with subsidies. "Made in India" chips is the new mantra.

What you fail to appriciate is that people do not buy semiconductors. They buy electronic goods. It is possible to have a massive electronic industry WITHOUT wasting money in setting up these white elephants called fabs.
Manufacture goods that sell. Manufacture PCBs, mechanical parts and finished goods. Once you have 100s of billions to burn, then establish your own top of the line fab industry. China did the same.

They did the same and then got sanctioned. So they decided to build their own fab plants now.

Why else do you think Huawei has now become a fab company?

Why can't Huawei import from Taiwan or Korea instead?

See, the thing is reality doesn't work the way you are assuming. As a big player, if you are not self-sufficient, you get cut out of the market by crook. I'm sure Huawei will run the fab plant at a loss and make it up in the final price of their smartphone.

The world isn't as fair as you think. The full chain needs to be in India. We will run it at a loss, and we won't export either. It's good as long as it's a product that cannot be sanctioned. The latest iPhone 12 Pro 128 GB costs $1100 in the US and $1600 in India. That's a $500 premium because we can't afford importing it.
 
By the time we "beat" Vietnam, BD, even China, we are gonna broke importing $200-300B worth of IC chips in 2030-35. We can't go to the IMF for a bail-out package just for ICs.

Thankfully, India is not going to follow what you said. GoI is pretty sure about the fab plants now. If we don't get access to the international market for the fab plant, then we cater to the domestic market only, even with subsidies. "Made in India" chips is the new mantra.



They did the same and then got sanctioned. So they decided to build their own fab plants now.

Why else do you think Huawei has now become a fab company?

Why can't Huawei import from Taiwan or Korea instead?

See, the thing is reality doesn't work the way you are assuming. As a big player, if you are not self-sufficient, you get cut out of the market by crook. I'm sure Huawei will run the fab plant at a loss and make it up in the final price of their smartphone.

The world isn't as fair as you think. The full chain needs to be in India. We will run it at a loss, and we won't export either. It's good as long as it's a product that cannot be sanctioned. The latest iPhone 12 Pro 128 GB costs $1100 in the US and $1600 in India. That's a $500 premium because we can't afford importing it.
It is like jet engine tech, nobody will share it. Build on your own or go broke importing it.

Long time back during 80's new zealand refused to export dairy cows to India! They were(still are) one of the biggest dairy products exporter and dint want to lose the market.
 
They did the same and then got sanctioned. So they decided to build their own fab plants now.
I think you misunderstood the sanction. Its not the sanction against the Fabs. Its sanction against IP and EDA tool.
Meaning, EVEN IF they have Fabs, they will run into this issue. Actually they have Fabs already in China YET they are facing this issue. Issue is, their flagship SoC platform Kirin uses ARM IP Core Licensed from ARM and they use EDA tools like Synopsys. To actually bypass this ban, they will need to have their own EDA toolchain (HARDER) and their own ARM core (HARD).

Source : New sanctions deal 'lethal blow' to Huawei. China decries US bullying

EVEN if they have their own fabs, they will STILL run in these issues. They wont be able to sell their products anywhere EVEN if they have their Fabs. No one will be able to buy chips from them or use their products without facing US sanctions.

Having a Fab in such a case will be a costly white elephant as you wont be able to utilize its capacity. The thing which you are missing is this : This is a technology denial ban and NOT a manufacturing denial ban.

For example, Huawei Ascend 910 chips are being fabricated in TSMC becauses it does not need any IP Cores from USA. All blocks come from Huawei itself.

1608876884795.png
 
Last edited:
I think you misunderstood the sanction. Its not the sanction against the Fabs. Its sanction against IP and EDA tool.
Meaning, EVEN IF they have Fabs, they will run into this issue. Actually they have Fabs already in China YET they are facing this issue. Issue is, their flagship SoC platform Kirin uses ARM IP Core Licensed from ARM and they use EDA tools like Synopsys. To actually bypass this ban, they will need to have their own EDA toolchain (HARDER) and their own ARM core (HARD).

Source : New sanctions deal 'lethal blow' to Huawei. China decries US bullying

EVEN if they have their own fabs, they will STILL run in these issues. They wont be able to sell their products anywhere EVEN if they have their Fabs. No one will be able to buy chips from them or use their products without facing US sanctions.

Having a Fab in such a case will be a costly white elephant as you wont be able to utilize its capacity. The thing which you are missing is this : This is a technology denial ban and NOT a manufacturing denial ban.

For example, Huawei Ascend 910 chips are being fabricated in TSMC becauses it does not need any IP Cores from USA. All blocks come from Huawei itself.

View attachment 18904

The Chinese have set up 3 new companies and poached talent from Synopsys and Cadence to make their own EDA tools. Also, Synopsys and Cadence are investing in these new companies because of what's at stake here for them. They are also making their own investments in China as well.

The biggest reason why people with skills in this segment are joining these companies is because the Chinese domestic market has matured to the point where it can sustain the entire semiconductor industry with its own market.

Annually, China consumes more than 50 percent of all semiconductors, both for internal use and eventual export. As such, the rapid growth of Chinese demand lifted the entire industry worldwide. However, domestic Chinese manufacturers are still only capable of meeting approximately 30 percent of their own demand.

daxue-consulting_Chinas-semiconductor-market_semiconductor-market-share-global-768x498.jpg


The Chinese already operate dozens of fab plants and they can't even meet 30% of their domestic demand. And here you guys are cribbing about 2 fab plants when India's population is expected to be at least 30% larger than China's over the next 2.5 decades.

Once silicon reaches its absolute hardware limit, we need to be ready with our own fabs that can take the world to the next generation otherwise the entire semiconductor industry will revolve around China in just a few years. If we start now, we will have the capacity to do it in 5-10 years.
 
It is like jet engine tech, nobody will share it. Build on your own or go broke importing it.

Long time back during 80's new zealand refused to export dairy cows to India! They were(still are) one of the biggest dairy products exporter and dint want to lose the market.

Yep. Pretty much the case here. China's already dominating the semiconductor market, it's only a matter of time before they dominate the entire industry given the new investments they are making under Made in China 2025. We need to create our own capacities to counter their rise. Lucky for us, their 2025 plan was pushed back due to the trade war, so that bought us some time. If we still don't utilise the time bought by the US, then that's gonna be one of the biggest blunders we ever make.