Indian Air Force: News & Discussions

[gryphus-scarface]

Many thanks. That would mean our govt should urgently work towards more awacs, better integration, better bvr missiles and it may be a great force multiplier. We can send different tejas aircraft to the front with agm to attack and others loaded with bvr to defend, both assisted by awacs or su30 radar.

Another question. Everyone says su30 is like mini awacs. Does it compensate for our lack of awacs aircraft?

The Su-30 has a very capable N011 Bars radar, but it simply isn't a replacement for an AWACS yet. It is a mini AWACS. We already have a decent number of AWACS. We supposedly purchased Airbus A330s to convert into AWACS. We also have the Astra and SFDR programmes for BVR missiles.

 

[gryphus-scarface]

We will replace everything from SPECTRA to EW to HMDS, cockpit etc. Since what is needed for flying is just airframe and FBW, changing electronics completely is possible. What problem is there if entire LRU and avionics of Rafale is replaced by the ones for Tejas? Can you explain?

Yes, the requirement for indigenisation is a demand and indigenisation means that India will be able to reproduce rafale fully. If France is unwilling to give technology to India, then it has to allow Indian owned technology to be substituted. That is why $4 billion of offset is paid. In fact 50% of the cost of deal is in offset! What offsets would anyone need to that extent?

This is not what indigenisation means. It simply means the parts will be Made in India. To what extent the parts will be made, depends on the deal. The idea that we will simply stick Uttam and get rid of SPECTRA is insane. That will have to be part of the deal. i.e. the deal will say that Uttam and other avionics will be used, as opposed to the Dassault avionics and Radar. That will be a part of India specific upgrades. As far as we know, no such requirements have been posed. And for the entire avionics to be replaced, that still needs certification. It is possible, but by no means trivial. The way this will work is, we buy Rafale, Dassault sets up a plant to manufacture 85% of planes. They will either make the radars here, or import the radars from France. Same with the EW suite. The chips will most likely be imported, unless they are fine with using SCL's 180nm node. Setting up a fab for a couple of planes is a massive waste of money.

 

[Advaidhya Tiwari]

As far as we know, no such requirements have been posed. And for the entire avionics to be replaced, that still needs certification.

How do you know what India specific upgrades were demanded? Your words show as if you know what was the deal despite it being called "secret" under secrecy clause!

Certification for avionics is not as hard or rigorous as certification for plane and FBW. Planes can fly and will not crash even if there is no avionics. Just engine, airframe and FBW is enough to make a plane fly safely

This is not what indigenisation means. It simply means the parts will be Made in India.

I don't want lessons in English. I am clearly speaking in context of Su30 being indigenous, the need for Rafale is non existent if it is not as indigenous as Su30. So, talk in the context, not out of it.

The way this will work is, we buy Rafale, Dassault sets up a plant to manufacture 85% of planes. They will either make the radars here, or import the radars from France. Same with the EW suite. The chips will most likely be imported, unless they are fine with using SCL's 180nm node. Setting up a fab for a couple of planes is a massive waste of money.

Most planes use chips upto 90nm as the high heat resistance needed for planes that can reach 250 celsius in heat while going supersonic needs to be handled. Lowe nm hips will get burnt out or short circuit. India already has 180nm chips fab and hence it makes no sense to import chips. India makes chips for Tejas inhouse too

Again, I am being very explicit about the use of non-indigenous rafale when India already has indigenous options. I can't understand how you continually refuse to answer that question and simply state your own monologue of 85% planes being made in India. The very point is that if the plane is not made fully in India without any import, then it is useless for Indian usage. So, there is simply no question of allowing anything to be imported.

Don't bother replying if you are not talking to the point

 

[gryphus-scarface]

How do you know what India specific upgrades were demanded? Your words show as if you know what was the deal despite it being called "secret" under secrecy clause!

You presume that the Rafale we buy will not have SPECTRA, its own EW suite, its own radar and IRST. At some point you have to realise its a very different Rafale. There's a reason our Su-30 MKI is very different from the Su-30 MKK or the Su-30SM.

Certification for avionics is not as hard or rigorous as certification for plane and FBW. Planes can fly and will not crash even if there is no avionics. Just engine, airframe and FBW is enough to make a plane fly safely

Right. We don't need our missiles to fire. We don't need to worry about how a glitch could make the airframe useless in combat. As long as it flies it is good enough. If only the IAF realised this earlier.

I don't want lessons in English. I am clearly speaking in context of Su30 being indigenous, the need for Rafale is non existent if it is not as indigenous as Su30. So, talk in the context, not out of it.

And do the Su-30s fly with Kaveri engine? Hint: they don't.

Most planes use chips upto 90nm as the high heat resistance needed for planes that can reach 250 celsius in heat while going supersonic needs to be handled. Lowe nm hips will get burnt out or short circuit. India already has 180nm chips fab and hence it makes no sense to import chips. India makes chips for Tejas inhouse too

You can't just change the node without having any impact on the chip's efficiency. I'm not even sure its possible. A new Fab would itself cost $6Billion (source : Reddit AMA of someone on the AJIT team). And the idea that lower node chips re less hear resistant is something I've never heard. The reason SCL never really was pushed so hard is because larger nodes are easier to protect from radiation in space.

Again, I am being very explicit about the use of non-indigenous Rafale when India already has indigenous options. I can't understand how you continually refuse to answer that question and simply state your own monologue of 85% planes being made in India. The very point is that if the plane is not made fully in India without any import, then it is useless for Indian usage. So, there is simply no question of allowing anything to be imported.

What you fail to understand is that indiginisation here means manufacturing here. Not using parts that have been designed here. Maybe in the future we will upgrade our Rafales with Uttam and a Kaveri derivative. We most definitely won't be using them as part of the original MMRCA bid.

Don't bother replying if you are not talking to the point

Give one source that even speculates Rafale using Uttam or Kaveri. Not just Rafale, any MMRCA contender. You are the only person who thinks that this is what indiginisation is. Don't bother replying without providing a source.

 

[Advaidhya Tiwari]

You presume that the Rafale we buy will not have SPECTRA, its own EW suite, its own radar and IRST. At some point you have to realise its a very different Rafale. There's a reason our Su-30 MKI is very different from the Su-30 MKK or the Su-30SM.

So what? That is the point. India just wants to reduce the time it takes to get a new plane designed and hence wants Rafale to use its Kaveri engine. Rafale's use of 75kN engine suits well with Kaveri's 81kN thrust. In past, Rafale was run on F404 too

Right. We don't need our missiles to fire. We don't need to worry about how a glitch could make the airframe useless in combat. As long as it flies it is good enough. If only the IAF realised this earlier.

Do you have mental problems? Why will a proven system which has been proven by trials fail just because it is used in Rafale? If UTTAM has been completed and proven to be used in MWF/MK1A, how will it misfire a missile in Rafale? What problem do you have?

And do the Su-30s fly with Kaveri engine? Hint: they don't.

Su30 flies with Al31FP made in HAL. So, it is still indigenous. No need of Kaveri as a result.

What you fail to understand is that indiginisation here means manufacturing here. Not using parts that have been designed here. Maybe in the future we will upgrade our Rafales with Uttam and a Kaveri derivative. We most definitely won't be using them as part of the original MMRCA bid.

Does it matter if we use it in initial rafale or final? I am only saying that by the time 150 rafales are delivered, India wll have ability to make Rafales indigenously without restriction in numbers & dependency

Give one source that even speculates Rafale using Uttam or Kaveri. Not just Rafale, any MMRCA contender. You are the only person who thinks that this is what indiginisation is. Don't bother replying without providing a source.

May be because I am the only reasonable person around. I have been the only person to speak of petroleum deals and how Arabs lowered oil price for Modi and this was eventually confirmed by Saudi oil minister in December 2018. I have been the only one to talk about many other extremely important things which turned out to be true. What matters is "logical consistency", not how many people are repeating what I am saying

 

[gryphus-scarface]

So what? That is the point. India just wants to reduce the time it takes to get a new plane designed and hence wants Rafale to use its Kaveri engine. Rafale's use of 75kN engine suits well with Kaveri's 81kN thrust. In past, Rafale was run on F404 too

Right. There won't be any problem with inlet and engine size. The fact that the airframe will have to be modified to fit the Kaveri is indeed irrelevant here. Obviously they'll do it for us at no extra cost.

Do you have mental problems? Why will a proven system which has been proven by trials fail just because it is used in Rafale? If UTTAM has been completed and proven to be used in MWF/MK1A, how will it misfire a missile in Rafale? What problem do you have?

It needs to integrate with everything else. We don't have a full EW suite like SPECTRA yet. That means it has to integrate with SPECTRA and other systems on the Rafale. The Rafale has sensor fusion, so it needs to fuse data from the Radar and IRST and other sensors (if any). But that won't be an issue since the Uttam was tested in a completely different condition for a completely different platform. We don't need to work for integration and recertification is a waste of time. We can just stick it in there.

Su30 flies with Al31FP made in HAL. So, it is still indigenous. No need of Kaveri as a result.

Exactly. Similary, Dassault will put up plants here for their engine. Alternatively, they will import, if the cost of setting up a plant is too high, they will import kits. See what Japan is doing with the F-35.

Does it matter if we use it in initial rafale or final? I am only saying that by the time 150 rafales are delivered, India wll have ability to make Rafales indigenously without restriction in numbers & dependency

May be because I am the only reasonable person around. I have been the only person to speak of petroleum deals and how Arabs lowered oil price for Modi and this was eventually confirmed by Saudi oil minister in December 2018. I have been the only one to talk about many other extremely important things which turned out to be true. What matters is "logical consistency", not how many people are repeating what I am saying

Delusions of granduer. You made one lucky guess. "logical consistency" is lacking in your arguments. You've quietly stopped talking about the chips that are needed for the Rafale.

 

[Advaidhya Tiwari]

Right. There won't be any problem with inlet and engine size. The fact that the airframe will have to be modified to fit the Kaveri is indeed irrelevant here. Obviously they'll do it for us at no extra cost.

The india specific modification and $4 billion offsets are the funds. Why act like retards when I have been saying this everytime? Also, the inlet size is sufficient for F404 means it is enough for Kaveri too. Simple common sense

It needs to integrate with everything else. We don't have a full EW suite like SPECTRA yet. That means it has to integrate with SPECTRA and other systems on the Rafale. The Rafale has sensor fusion, so it needs to fuse data from the Radar and IRST and other sensors (if any). But that won't be an issue since the Uttam was tested in a completely different condition for a completely different platform. We don't need to work for integration and recertification is a waste of time. We can just stick it in there.

Again, talking nonsense. India is developing EW suite and is in advanced stage. We have already installed our EW in Jaguar & few other planes. We are developing a more capable one for Tejas which will be sued in Indian rafale too.

The sensor fusion technology is just about integration, not magic. These fusion is existent in every modern plane in one form or another. Even Tejas will have sensor fusion.

We will need to integrate new electronics in Rafale and certify it but it will be an easy process. Radar is a radar, is used from nose cone of Tejas or Rafale. So, the idea that it will be hard to take it and use it in another plane is foolishness. There will be some changes needed but that will be minor enough. Just like Israel radar can be used in Tejas despite Israel not developing it for Tejas, UTTAM can be used in Rafale. Why don't you use analogy instead of arrogance?

Exactly. Similary, Dassault will put up plants here for their engine. Alternatively, they will import, if the cost of setting up a plant is too high, they will import kits. See what Japan is doing with the F-35.

Either Dassault will give full TOT like Al31FP or it will use Kaveri. Japan importing F35 kits is not to be compared with India. India will not accept any kits. Either Dassault gets it 100% mad in India without dependency or we don't need Rafale.

Delusions of granduer. You made one lucky guess. "logical consistency" is lacking in your arguments. You've quietly stopped talking about the chips that are needed for the Rafale.

One lucky guess? I have been telling everything in a very consistent and logical manner and answering every question posed in best possible manner. That is no lucky guess. I knew what I was saying.

I have not stopped about chips. I have clarified that India can make chips for defence needs. India has 180nm fab in SCL, Chandigarh which is optimal for defence equipment. Defence equipment cant use the latest chips used in mobile phones as the latest generation is not heat resistant. So, India has all the chip manufacturing from Silicon semiconductor to GaAs semiconductor made in India. GaAs is made in GAETEC and microprocessor is made in SCL.

 

[gryphus-scarface]

The india specific modification and $4 billion offsets are the funds. Why act like retards when I have been saying this everytime? Also, the inlet size is sufficient for F404 means it is enough for Kaveri too. Simple common sense


Again, talking nonsense. India is developing EW suite and is in advanced stage. We have already installed our EW in Jaguar & few other planes. We are developing a more capable one for Tejas which will be sued in Indian rafale too.

Yes the jey word is developing. It is not ready yet.

The sensor fusion technology is just about integration, not magic. These fusion is existent in every modern plane in one form or another. Even Tejas will have sensor fusion.

Here let me explain it this way. Uttam will present the data it gets in some way. What will the Rafale sensor fusion software expect? The data format that the Rafale's original radar gave. How will it change he modes on the Uttam radar? How will it interpret data which looks like nonesense to it? This interface is not black magic, but it is important. And without knowing how to interface, how will we integrate? Your counter argument is we will replace everything.

We will need to integrate new electronics in Rafale and certify it but it will be an easy process. Radar is a radar, is used from nose cone of Tejas or Rafale. So, the idea that it will be hard to take it and use it in another plane is foolishness. There will be some changes needed but that will be minor enough. Just like Israel radar can be used in Tejas despite Israel not developing it for Tejas, UTTAM can be used in Rafale. Why don't you use analogy instead of arrogance?

Easy? It takes time and effort. What is easy is manufacturing the existing certified and tested stuff that makes up the Rafales that are flying today. And you still claim that radars are interchangeable.

Either Dassault will give full TOT like Al31FP or it will use Kaveri. Japan importing F35 kits is not to be compared with India. India will not accept any kits. Either Dassault gets it 100% mad in India without dependency or we don't need Rafale.

Dassault will set up a plant in India or import kits. It depends on the cost.

One lucky guess? I have been telling everything in a very consistent and logical manner and answering every question posed in best possible manner. That is no lucky guess. I knew what I was saying.

Yes. And your logic is that Indian variants are available so Indian things will be used. Why not just use the Tejas while we're at it? It is more Indian than a Rafale will ever be. The airframe was made here. Rafale airframe was designed in France. Why just build the plane here?

I have not stopped about chips. I have clarified that India can make chips for defence needs. India has 180nm fab in SCL, Chandigarh which is optimal for defence equipment. Defence equipment cant use the latest chips used in mobile phones as the latest generation is not heat resistant. So, India has all the chip manufacturing from Silicon semiconductor to GaAs semiconductor made in India. GaAs is made in GAETEC and microprocessor is made in SCL.

Right. 180nm is ancient history now. By no means recent. The k8 family from AMD used 130nm in 2001, and not too soon after shifted to smaller nodes. That means 180nm is a full 20 years behind. And you claim heat resistance is an issue. I've yet to see any chip from any node survive high temperatures. They will all melt the same way. Node has no relevance. Is this your logical argument? The node only helps improve efficiency, by reducing transistor size, and thus the current needed to keep each transistor / electric component in its respective state. The only other thing that is affected is density, so you can make the same chip smaller. Regardless of the chip used they will all need some sort of cooling, passive or active. 180nm has no bearing on this. So Dassault will need to import, unless someone sets up a fab with node size that they need.

 

[Advaidhya Tiwari]

Yes the jey word is developing. It is not ready yet.

Ok, my bad. The word I must have used is "upgrading" as Jaguar already has Indian EW suite. It is just an upgrade for modern times

Here let me explain it this way. Uttam will present the data it gets in some way. What will the Rafale sensor fusion software expect? The data format that the Rafale's original radar gave. How will it change he modes on the Uttam radar? How will it interpret data which looks like nonesense to it? This interface is not black magic, but it is important. And without knowing how to interface, how will we integrate? Your counter argument is we will replace everything.

That is because entire set of electronics including cables, cockput will be replaced by Indian ones which are going in Tejas

Easy? It takes time and effort. What is easy is manufacturing the existing certified and tested stuff that makes up the Rafales that are flying today. And you still claim that radars are interchangeable.

The ones that go in Tejas are also tested. Think about using a music deck in Omni van vs i10 car, does it matter? Same way, the avionics and electronics are independent of airframe. So, the new electronics can be added. It is just that one item can't be switched but entire items have to be switched.

Dassault will set up a plant in India or import kits. It depends on the cost.

Who is talking of cost here? It is about technology acquisition Knowledge is not measured in cost.

Yes. And your logic is that Indian variants are available so Indian things will be used. Why not just use the Tejas while we're at it? It is more Indian than a Rafale will ever be. The airframe was made here. Rafale airframe was designed in France. Why just build the plane here?

Tejas is too small and the Kaveri made for Tejas is now useless as Tejas MK1 is being cut down t just 123 order. So, by getting Rafale design, we will have Indian plane which will make up for the lost time in developing a underpowered Tejas.

Right. 180nm is ancient history now. By no means recent. The k8 family from AMD used 130nm in 2001, and not too soon after shifted to smaller nodes. That means 180nm is a full 20 years behind. And you claim heat resistance is an issue. I've yet to see any chip from any node survive high temperatures. They will all melt the same way. Node has no relevance. Is this your logical argument? The node only helps improve efficiency, by reducing transistor size, and thus the current needed to keep each transistor / electric component in its respective state. The only other thing that is affected is density, so you can make the same chip smaller. Regardless of the chip used they will all need some sort of cooling, passive or active. 180nm has no bearing on this. So Dassault will need to import, unless someone sets up a fab with node size that they need.

180 nm was the technology till 2001. 130nm started in 2002. But the latest technology ha sstagnated at 14 nm (intel is still dveeloping 10nm and the 10nm of Taiwan etc are 14nm being falsely tagged at 10nm). So, due to the saturation level, the ones closest to the saturation level is not used where reliability is needed.

Military items are not interested in running on batteries or power saving as in mobile or desktop usage. What matters more is robustness and reliability. Every chip needs some form of cooling but the smaller nodes require extreme amount of cooling and can't handle even minor fluctuations.

The chips, whether of 180nm or 14nm are made of same material and hence have same conductivity and "leakage" current. This conductivity and leakage current increase as temperature increase. As the conductivity and leakage current increase beyond a point, nodes will get shorted. Here the size of nodes matter. Higher size of the node will mean that the larger ones will have more resistance and much higher short-circuit current requirement. This on one hand increased power usage but on the other hand reduces susceptibility to changes in temperature.

The reason why GaN is considered better than GaAs is also due to the higher temperature tolerance of GaN which drastically increase performance. It is not about power efficiency of GaN itself but the reduction in cooling requirement that is what matters.

That is why only those items which need high power efficiency uses latest generation semiconductor. In planes, the focus is more on heat resistance and robustness, not power efficiency. Even a small glitch can mean that the plane will crash. Whereas the plane generates power in several kilowatts which is more than enough to power older generation processors. There is no high end graphics or games to play either and hence the 180nm processors are enough for processing the limited data points.

Even planes like F22 use outdated processors from 1990s. F35 also uses older processors of early 2000s. Rafale would also be using the older generation chipset like F22. The state of the art military items like missiles, space rockets etc generally use processors of 1um (1000nm), 800nm or of other 1990s generation.

India has enough technology in terms of semiconductors for defence applications. There is no need to import chips. That is also how India is making chips for Tejas indigenously. Dassault will not need to import anything for Rafale indigenisation

 

[garg_bharat]

I have always maintained that the F-18 is the only viable American aircraft that India can consider if it really has to buy American. However, I think this is hogwash. Rafale for the IAF is a done deal. So then purchasing SH for IN does not make any sense logistically as well as financially.

No, very likely India will buy American.
There are many reasons.

1. India has a trade surplus with USA. This has become a bone of contention in USA. Buying more American weapons balances the trade to some extent plus gives India leverage in the West.

2. India has a trade deficit with EU in comparison. So buying from EU does not help in balancing trade.

3. There is an increasing realization that India needs USA in case a war with China breaks out. Russia no longer has the spare capacity and politically Russia is trapped. Even Europe has little excess capacity. Only USA has a military industrial infrastructure to support big wars.

4. Our most profitable industries - software and BPO - are irreversibly tied to US economy.

 

[gryphus-scarface]

Ok, my bad. The word I must have used is "upgrading" as Jaguar already has Indian EW suite. It is just an upgrade for modern times


That is because entire set of electronics including cables, cockput will be replaced by Indian ones which are going in Tejas


The ones that go in Tejas are also tested. Think about using a music deck in Omni van vs i10 car, does it matter? Same way, the avionics and electronics are independent of airframe. So, the new electronics can be added. It is just that one item can't be switched but entire items have to be switched.

The question is why will they do this when it is more effort? Your entire logic is that it is less effort to do so, and thus they will integrate Uttam, the EW suite, and other avionics. This is generally done when the indigenous options are better. As of now, we can only say that the indigenous options are at the same level as Rafale at best. Note that our Advanced Self Protection Jammer is still in very early phase of development. Our IRST will similarly take time. We still don't have anything like SPECTRA's active cancellation.

Who is talking of cost here? It is about technology acquisition Knowledge is not measured in cost.

A single chip fab can run into billions of dollars to set up. Why do you think we have only the SCL? According to someone on the AJIT team upgrading SCL to get to 90nm will cost roughly $6Billion. Now think why would Dassault set up a chip fab here? That too for less than 100 planes, for roughly 300 chips max?

180 nm was the technology till 2001. 130nm started in 2002. But the latest technology ha sstagnated at 14 nm (intel is still dveeloping 10nm and the 10nm of Taiwan etc are 14nm being falsely tagged at 10nm). So, due to the saturation level, the ones closest to the saturation level is not used where reliability is needed.

Military items are not interested in running on batteries or power saving as in mobile or desktop usage. What matters more is robustness and reliability. Every chip needs some form of cooling but the smaller nodes require extreme amount of cooling and can't handle even minor fluctuations.

The chips, whether of 180nm or 14nm are made of same material and hence have same conductivity and "leakage" current. This conductivity and leakage current increase as temperature increase. As the conductivity and leakage current increase beyond a point, nodes will get shorted. Here the size of nodes matter. Higher size of the node will mean that the larger ones will have more resistance and much higher short-circuit current requirement. This on one hand increased power usage but on the other hand reduces susceptibility to changes in temperature.

The reason why GaN is considered better than GaAs is also due to the higher temperature tolerance of GaN which drastically increase performance. It is not about power efficiency of GaN itself but the reduction in cooling requirement that is what matters.

That is why only those items which need high power efficiency uses latest generation semiconductor. In planes, the focus is more on heat resistance and robustness, not power efficiency. Even a small glitch can mean that the plane will crash. Whereas the plane generates power in several kilowatts which is more than enough to power older generation processors. There is no high end graphics or games to play either and hence the 180nm processors are enough for processing the limited data points.

Even planes like F22 use outdated processors from 1990s. F35 also uses older processors of early 2000s. Rafale would also be using the older generation chipset like F22. The state of the art military items like missiles, space rockets etc generally use processors of 1um (1000nm), 800nm or of other 1990s generation.

India has enough technology in terms of semiconductors for defence applications. There is no need to import chips. That is also how India is making chips for Tejas indigenously. Dassault will not need to import anything for Rafale indigenisation

There's more to reduced node size than power saving. Intel had 130nm by 2002, and 90nm by 2004. You claim that the F-22 flies with CPUs of 90s, but it actually doesn't. The original lots did. They flew with an Intel i960MX. The later lots are flying with a PowerPC G5 based CPU (Raytheon CIP). That's much more contemporary. Also note that the F-22's CIP is liquid cooled. http://www.f-22raptor.com/news_view.php?nid=15&yr=2000

The main advantage of a smaller node for military is that less heat is produced. And the same number of transistors can be put in a much smaller chip.

I don't know where you've got the idea that space stuff uses 1um. SCL was set up for ISRO. And the reason they use 180nm is because of radiation hardening. A larger transistor size implies that each transistor is less susceptible to changing its charge levels due to radiation. For e.g. is there's some alpha rays passing through the microprocessor then they will be less likely to lose their charge and enter an invalid state. Not because of cooling issues.

One of other problems with the F-22 was that the original batches used an outdated intel processor which was discontinued. Nobody will set up older nodes for you. We have an older node, but making sure that Dassault's chips can be made here isn't so easy.

You might like this RFI for MMRCA: http://indianairforce.nic.in/sites/default/files/RFI_Fighter_Aircraft.pdf

You'll see that IAF is asking for "Indigenisation content", under part 2. It specifically asks what level of indigenisation will be achieved.

at no point do they say, please allow us to integrate our own avioinics, radar, microprocessor, etc.

 

[Advaidhya Tiwari]

The question is why will they do this when it is more effort? Your entire logic is that it is less effort to do so, and thus they will integrate Uttam, the EW suite, and other avionics. This is generally done when the indigenous options are better. As of now, we can only say that the indigenous options are at the same level as Rafale at best. Note that our Advanced Self Protection Jammer is still in very early phase of development. Our IRST will similarly take time. We still don't have anything like SPECTRA's active cancellation.

Being same level means there is no reason to get foreign items. Isn't it common sense that it is less effort to use indigenous items in the long run? Over time, there will be upgrades which will be 100 times easier on indigenous items than imports. So, long term plan says that indigenous ones are much better in all ways.

Our SPJ & IRST will be completed before MWF becomes FOC in 2025. We already have RWR & SPJ pod made for Su30 MKI which only needs upgrade. IRST is something we have to make from scratch but we have acquired several IIR technology which makes a solid base for it.

A single chip fab can run into billions of dollars to set up. Why do you think we have only the SCL? According to someone on the AJIT team upgrading SCL to get to 90nm will cost roughly $6Billion. Now think why would Dassault set up a chip fab here? That too for less than 100 planes, for roughly 300 chips max?

We can set up 14-22nm fab for less than 6 billion dollars! It won't take $6 billion to upgrade SCL to 90nm level unless it is imported and the foreign company charge high royalty.

None said that Dassault will set up a chip fab. I am only syaing that it will not be needed as India can make the chips for planes in India itself.

There's more to reduced node size than power saving. Intel had 130nm by 2002, and 90nm by 2004. You claim that the F-22 flies with CPUs of 90s, but it actually doesn't. The original lots did. They flew with an Intel i960MX. The later lots are flying with a PowerPC G5 based CPU (Raytheon CIP). That's much more contemporary. Also note that the F-22's CIP is liquid cooled. http://www.f-22raptor.com/news_view.php?nid=15&yr=2000

The main advantage of a smaller node for military is that less heat is produced. And the same number of transistors can be put in a much smaller chip.

I don't know where you've got the idea that space stuff uses 1um. SCL was set up for ISRO. And the reason they use 180nm is because of radiation hardening. A larger transistor size implies that each transistor is less susceptible to changing its charge levels due to radiation. For e.g. is there's some alpha rays passing through the microprocessor then they will be less likely to lose their charge and enter an invalid state. Not because of cooling issues.

One of other problems with the F-22 was that the original batches used an outdated intel processor which was discontinued. Nobody will set up older nodes for you. We have an older node, but making sure that Dassault's chips can be made here isn't so easy.

You might like this RFI for MMRCA: http://indianairforce.nic.in/sites/default/files/RFI_Fighter_Aircraft.pdf

You'll see that IAF is asking for "Indigenisation content", under part 2. It specifically asks what level of indigenisation will be achieved.

at no point do they say, please allow us to integrate our own avioinics, radar, microprocessor, etc.

What is the lithography of PowerPC G5? How many nano metre? You gave me a link of year 2000 when the 180nm had not yet arrived. So, F22 does use 1990s processor!

India has 800nm & 1um processor fab in BEL, SITAR etc to be used in missiles and other state of the art technology. ISRO uses a variety of chips according to the conditions suitable.

The main advantage of smaller node is that power consumption is less and hence heat generation is also less. But the bigger problem is when there is heat in external environment. The heat generated when a plane goes in 1.5mach is 200-250 celsius. This is a major reason for problems. Internal heat is actually lesser in smaller node and hence it is preferred in civilian usage. But when high stress condition is there, then then smaller nodes fail. As I said, we don't lack power in planes to power 2 higher node processors instead of one small node processor. Also, there is no reason to demand high transistor count. The utility of processors in a plane is limited. Unlike your laptop, it does not have to run 3D graphics or video games and files of 20GB in size!

About RFI for MMRCA, it is called RFI for a reason. It seeks out information. Only in the finalised deal will there be the conditions. Moreover, did you ever ask the reason for RFI for MMRCA2 when Rafale deal involved so much offsets which compel India to buy Rafale once again? Already the offset of $4 billion for 36 rafales cover most of the indigenisation conditions & is kept under secrecy clause.

 

[gryphus-scarface]

Being same level means there is no reason to get foreign items. Isn't it common sense that it is less effort to use indigenous items in the long run? Over time, there will be upgrades which will be 100 times easier on indigenous items than imports. So, long term plan says that indigenous ones are much better in all ways.

Our SPJ & IRST will be completed before MWF becomes FOC in 2025. We already have RWR & SPJ pod made for Su30 MKI which only needs upgrade. IRST is something we have to make from scratch but we have acquired several IIR technology which makes a solid base for it.

So lets hold of buying Rafale or any other plane until our own system is ready. In fact, if we wiat another 10 years, AMCA will be ready, so we can just order more AMCAs. No need for Rafale. AMCA will be more indigenous than Rafale will ever be.

We can set up 14-22nm fab for less than 6 billion dollars! It won't take $6 billion to upgrade SCL to 90nm level unless it is imported and the foreign company charge high royalty.

Who will set it up for so cheap? You do realise that there is a lot of IP and very precise engineering to make the fabrication plants right? Each transistor is a couple of nanometres across. That needs to cut and maintained within a very small margin of error. Nobody will set that up for cheap. India will have to develop on its own, and even then the development cost is very high. Ask Intel, TSMC, Samsung, etc. about the costs involved. This has been confirmed by the AJIT tema. I think they have more expertise than both you and me in this.

None said that Dassault will set up a chip fab. I am only syaing that it will not be needed as India can make the chips for planes in India itself.

And who will give us the design to fabricate? For now it reamins much cheaper for Dassault to dimply get their own chips from wherever they manufacture Rafales.

What is the lithography of PowerPC G5? How many nano metre? You gave me a link of year 2000 when the 180nm had not yet arrived. So, F22 does use 1990s processor!

Best I can find is 130nm to 90nm. Still ahead of SCL. And the PowerPC G5 is by no means 90s tech. It is early 2000s.

India has 800nm & 1um processor fab in BEL, SITAR etc to be used in missiles and other state of the art technology. ISRO uses a variety of chips according to the conditions suitable.

They having outdating fabrication plants has nothing to do with necessity. Nobody in the world maintains such outdated plants.

The main advantage of smaller node is that power consumption is less and hence heat generation is also less. But the bigger problem is when there is heat in external environment. The heat generated when a plane goes in 1.5mach is 200-250 celsius. This is a major reason for problems. Internal heat is actually lesser in smaller node and hence it is preferred in civilian usage. But when high stress condition is there, then then smaller nodes fail. As I said, we don't lack power in planes to power 2 higher node processors instead of one small node processor. Also, there is no reason to demand high transistor count. The utility of processors in a plane is limited. Unlike your laptop, it does not have to run 3D graphics or video games and files of 20GB in size!

Yes but it needs to keep of track of many targets, communicate with other planes, and do a whole host of other things, while leaving headroom for future improvements. The problem isn't that of power generation. You keep shifting the problem. These chips won't melt any more easily than the chips on higher nodes. Again, these chips are liquid cooled. I have shown you the link for the Raptor. It can keep the chips cool enough. Please stop saying that the ships will burn at high speeds. They are not being air cooled.

About RFI for MMRCA, it is called RFI for a reason. It seeks out information. Only in the finalised deal will there be the conditions. Moreover, did you ever ask the reason for RFI for MMRCA2 when Rafale deal involved so much offsets which compel India to buy Rafale once again? Already the offset of $4 billion for 36 rafales cover most of the indigenisation conditions & is kept under secrecy clause.

The 36 Rafales are being made in France with no indigenous technology. They are pretty much the same plane that Qatar has. The offset is where Dassault and Thales decide to reinvest in India. Chosen parters include Reliance and DRDO. nothing to do with the MMRCA deal.

 

[IndianHawk]

To have a better understanding of what iaf can do in future one needs to have a look at iaf capita budget and see the growth in that allocation over past ten years. Then project iaf capital budgets for next 10 year.
Then we can have a rough estimate of how much money iaf can spend of capital acquisitions over next decade.
And finally we can fit rafale , mwf , mk1a , amca , su30 upgrade into that budget to see what is feasible .
Does anybody has data.

Sent from my C103 using Tapatalk

 

[Advaidhya Tiwari]

So lets hold of buying Rafale or any other plane until our own system is ready. In fact, if we wiat another 10 years, AMCA will be ready, so we can just order more AMCAs. No need for Rafale. AMCA will be more indigenous than Rafale will ever be

Actually that was the plan. Since we make Su30 anyways, why do we need Rafale? Parrikar also asked the same question. Then France loosened up and gave better deal which is most likely what I said - UTTAM, Kaveri K10, DARE EW and HAL-SAMTEL avionics on Rafale, making it fully Indian. This will give India another plane airframe design of rafale and give more variety for usage

Who will set it up for so cheap? You do realise that there is a lot of IP and very precise engineering to make the fabrication plants right? Each transistor is a couple of nanometres across. That needs to cut and maintained within a very small margin of error. Nobody will set that up for cheap. India will have to develop on its own, and even then the development cost is very high. Ask Intel, TSMC, Samsung, etc. about the costs involved. This has been confirmed by the AJIT tema. I think they have more expertise than both you and me in this.

We are already making 180nm isn't it? hat also requires precision engineering. The $6 billion was the cost for making fab for 22nm as proposed by AMD with HSL, Cricket semiconductor etc. Costing $6 billion for upgrading 180nm to 90nm is a bit too much. But it is irrelevant to topic and hence I won't argue. In any case, France is not willing to give any semiconductor technology to India for any price.

And who will give us the design to fabricate? For now it reamins much cheaper for Dassault to dimply get their own chips from wherever they manufacture Rafales.

This is a valid question. But the qustion I am asking is whether we need Dassault chips if we change entire electronics to Indian ones? Won't that mean that India will also be using Indian chips and hence have the design?

Best I can find is 130nm to 90nm. Still ahead of SCL. And the PowerPC G5 is by no means 90s tech. It is early 2000s.

PowerPC G5 has many versions and is a series like intel i5. There are 2nd gen i5, 4th gen i5 etc. Be assured, F22 has chips of 2000 or earlier technology and hence have chips of 180nm or less.

They having outdating fabrication plants has nothing to do with necessity. Nobody in the world maintains such outdated plants.

As a matter of fact, USA has outdated fab just to service its strategic needs. We keep defence and civilian plants separately for a reason. The defence plants use proven technology and not fancy modern technology. Even if the fab lithography technology is suitable, it will take about 10 years from introduction for the defence industry to use it as it takes time to solve all errors and other problems and then create a design architecture for use

Yes but it needs to keep of track of many targets, communicate with other planes, and do a whole host of other things, while leaving headroom for future improvements. The problem isn't that of power generation. You keep shifting the problem. These chips won't melt any more easily than the chips on higher nodes. Again, these chips are liquid cooled. I have shown you the link for the Raptor. It can keep the chips cool enough. Please stop saying that the ships will burn at high speeds. They are not being air cooled.

I know that they are liquid cooled but even then cooling needs will be much higher when external environment is stressful. The power can minutely fluctuate as the plane takes a high G turn, there can be other problems too in such stressful environment. The chips will not melt. Melting is never the problem. The problem is with the chips shorting between transistors and then causing bugs in processing or simply shutting down.

If you had a PC bought in early 2000, You could see that GTA Vice City type games could be played on 180nm desktops. The 180nm chips are no joke and not to be underestimated. GTA VC released in 2002 and was developed for 180nm processor itself. I myself have played it on 130nm desktop and it worked flawlessly.

You are exaggerating the need for high speed processing in planes or radars. It is true that high speed processing is desirable but that is within limits. We don't need ultra high speed processing. The processors are specialised in architecture and designed for specific tasks. Multiple processors are kept for multiple functions. So, there is no scope for "improvements". If you want a new feature, you must add a new processor. Unlike a computer, you are not having a general processor which can do mathematics, render videos, play and synthesise dynamic graphics etc. So, the speeds offered by 180nm processors is more than enough to do all the jobs and yet be well below the threshold limit.

However, keeping smaller node processor can increase cooling needs without adding any benefits. It also reduces reliability in terms of short circuiting. Processors upto 90nm may be used without much risks but beyond 90nm, there will be considerable increase in reliability risks.

The 36 Rafales are being made in France with no indigenous technology. They are pretty much the same plane that Qatar has. The offset is where Dassault and Thales decide to reinvest in India. Chosen parters include Reliance and DRDO. nothing to do with the MMRCA deal.

That is an assumption. The $4 billion offset deal is the one in discussion. $4 billion out of $9 billion is offset while only $4.5 billion is for Rafale. Govt has stated that each Rafale cost Rs 670 crore. this means that 36 Rafales must have cost 24000 crore. Another 700 million was maintenance and spare parts for 5 years. But the total deal was for 59000 crore. So, where did 30000 crore go? That is also the main reason of Congress trying to call it as "corruption". This offset was kept as "secret" in the deal. Companies like DRAL only got Rs 1000 crore and hence does not amount much. Where is the rest of the offset?

So, clearly, the MMRCA2 also has been decided beforehand. But India wanted to "trust after verification" and hence held back te entire deal by splitting it in 2 parts

 

[gryphus-scarface]

Actually that was the plan. Since we make Su30 anyways, why do we need Rafale? Parrikar also asked the same question. Then France loosened up and gave better deal which is most likely what I said - UTTAM, Kaveri K10, DARE EW and HAL-SAMTEL avionics on Rafale, making it fully Indian. This will give India another plane airframe design of rafale and give more variety for usage


We are already making 180nm isn't it? hat also requires precision engineering. The $6 billion was the cost for making fab for 22nm as proposed by AMD with HSL, Cricket semiconductor etc. Costing $6 billion for upgrading 180nm to 90nm is a bit too much. But it is irrelevant to topic and hence I won't argue. In any case, France is not willing to give any semiconductor technology to India for any price.

Its because the cost is exponential. You seem to think that the cost for upgrading an existing fab isn't high. It is very high. Its why we don't have any private fabs around. Its also why new fabs aren't set up willy nilly. In fact last year's declining PC sales were traced back to Intel's inability to produce chips fast enough. The world leaders in fabrication couldn't set up enough capital for a new node + fab. How do you think that it is easy to upgrade?
Complexity rises exponentially. That is why fabs are a big deal and so costly.

This is a valid question. But the qustion I am asking is whether we need Dassault chips if we change entire electronics to Indian ones? Won't that mean that India will also be using Indian chips and hence have the design?

And the question I'm asking is : do we have the capabilities needed? The Rafale has a whole host of sensors and other tech. We need to replace all of that, along with the microprocessor, and the

PowerPC G5 has many versions and is a series like intel i5. There are 2nd gen i5, 4th gen i5 etc. Be assured, F22 has chips of 2000 or earlier technology and hence have chips of 180nm or less.

Yes but first introduced in 2002/2003, with production until 2005. It had releases in 130nm and 90nm. I'm not sure which variant the Raptor has, but it was the latest available at the time, since Raptor obtained FOC around then.

As a matter of fact, USA has outdated fab just to service its strategic needs. We keep defence and civilian plants separately for a reason. The defence plants use proven technology and not fancy modern technology. Even if the fab lithography technology is suitable, it will take about 10 years from introduction for the defence industry to use it as it takes time to solve all errors and other problems and then create a design architecture for use

I'd like to see some concrete info on these fabs. 1um sounds insane.

I know that they are liquid cooled but even then cooling needs will be much higher when external environment is stressful. The power can minutely fluctuate as the plane takes a high G turn, there can be other problems too in such stressful environment. The chips will not melt. Melting is never the problem. The problem is with the chips shorting between transistors and then causing bugs in processing or simply shutting down.

You keep saying stressful environment. Please give exact details. High g turns won't affect newer node chips any more than older node chips. Power fluctuations will affect older nodes just as much. Shorting between transistors on smaller nodes is something I've never heard of. Shorting happens when voltage is high enough between two points. If you have any source for this claim please provide it.

If you had a PC bought in early 2000, You could see that GTA Vice City type games could be played on 180nm desktops. The 180nm chips are no joke and not to be underestimated. GTA VC released in 2002 and was developed for 180nm processor itself. I myself have played it on 130nm desktop and it worked flawlessly.

You are exaggerating the need for high speed processing in planes or radars. It is true that high speed processing is desirable but that is within limits. We don't need ultra high speed processing. The processors are specialised in architecture and designed for specific tasks. Multiple processors are kept for multiple functions. So, there is no scope for "improvements". If you want a new feature, you must add a new processor. Unlike a computer, you are not having a general processor which can do mathematics, render videos, play and synthesise dynamic graphics etc. So, the speeds offered by 180nm processors is more than enough to do all the jobs and yet be well below the threshold limit.

However, keeping smaller node processor can increase cooling needs without adding any benefits. It also reduces reliability in terms of short circuiting. Processors upto 90nm may be used without much risks but beyond 90nm, there will be considerable increase in reliability risks.

Its not a question of processing power for today's needs. Its a question of tomorow's needs. What happens when we come up with an unmanned wingman? What about other cases we haven't even thought of? And no game is ever released for some node. It is released for specific hardware. The benefit of smaller nodes are clear. It allows us to make smaller cheaps that need less power, and generate less heat. Less heat means smaller cooling system is sufficient. Smaller cooling system means more space for other stuff. The LCA already has issues with space.

That is an assumption. The $4 billion offset deal is the one in discussion. $4 billion out of $9 billion is offset while only $4.5 billion is for Rafale. Govt has stated that each Rafale cost Rs 670 crore. this means that 36 Rafales must have cost 24000 crore. Another 700 million was maintenance and spare parts for 5 years. But the total deal was for 59000 crore. So, where did 30000 crore go? That is also the main reason of Congress trying to call it as "corruption". This offset was kept as "secret" in the deal. Companies like DRAL only got Rs 1000 crore and hence does not amount much. Where is the rest of the offset?

So, clearly, the MMRCA2 also has been decided beforehand. But India wanted to "trust after verification" and hence held back te entire deal by splitting it in 2 parts

Reliance got the largest share. That is why people were complaining. Not because the money went away. 50% was reinvested in India. The rest goes to Dassault. Not too hard to understand.
The most was given to DRAL. Smaller amounts were given to other companies like BTSL, DEFSYS, Kinetic, Mahindra, Maini, SAMTEL, etc. https://www.dassault-aviation.com/e...tract-india-clarifications-dassault-aviation/

https://www.livefistdefence.com/201...of-frances-e4-billion-india-offsets-plan.html

https://www.thehindu.com/news/natio...-agreements/article26775545.ece?homepage=true

 

[Advaidhya Tiwari]

Its because the cost is exponential. You seem to think that the cost for upgrading an existing fab isn't high. It is very high. Its why we don't have any private fabs around. Its also why new fabs aren't set up willy nilly. In fact last year's declining PC sales were traced back to Intel's inability to produce chips fast enough. The world leaders in fabrication couldn't set up enough capital for a new node + fab. How do you think that it is easy to upgrade?
Complexity rises exponentially. That is why fabs are a big deal and so costly.

Then how India upgraded into 180nm? Intel has enough fab facilities. In fact, intel does not make chips for mobile phones. SO, the idea that intel could not make enough chips for PC is ridiculous. If intel could make 100 chips last year and this year, the sales are down to 85 means 15% capacity is excess, not reduced. Since intel does not make mobile chips, there is no excuse for not being able to make chips for PCs

And the question I'm asking is : do we have the capabilities needed? The Rafale has a whole host of sensors and other tech. We need to replace all of that, along with the microprocessor, and the

By the time the Rafale indigenisation sets in (after 36 rafales delivered in 2023), India will be having the capability needed

You keep saying stressful environment. Please give exact details. High g turns won't affect newer node chips any more than older node chips. Power fluctuations will affect older nodes just as much. Shorting between transistors on smaller nodes is something I've never heard of. Shorting happens when voltage is high enough between two points. If you have any source for this claim please provide it

Stressful environment means:
1) External temperature going to 200 celsius. So, a chip must be made to endure 300 celsius temperature as a safety measure
2) The power supply is not some stable powerplant but the engine exhaust. So, minor fluctuations are expected
3) Potential surge current when the plane gets hit by some missile or a bird

In fact, shorting between nodes is one of the most common thing. Most processors have redundancy built in due to which the processor diverts the processing to some other node when this happens. All processors have some defective nodes but the redundancy takes care of that. As the nodes get smaller, the chances of defects increase. This is further exacerbated if external factors like temperature rises.

Shorting happens due to:
1) defects in fabrication which in turn is higher in smaller nodes as the precision required is high
2) usage causing the nodes to wear out
3) temperature fluctuation causing the conductivity to rise and hence the leakage current to saturate

The voltage being high enough is depends on what is "enough". When temperature rises, due to conductivity rise, the voltage requirement will reduce. So, what voltage worked in room temperature may not work. Also, the holes and electrons can move around faster when temperature increases which adds to degradation. All this is pronounced when the margin of error is very thin due to thin nodes.

Its not a question of processing power for today's needs. Its a question of tomorow's needs. What happens when we come up with an unmanned wingman? What about other cases we haven't even thought of? And no game is ever released for some node. It is released for specific hardware. The benefit of smaller nodes are clear. It allows us to make smaller cheaps that need less power, and generate less heat. Less heat means smaller cooling system is sufficient. Smaller cooling system means more space for other stuff. The LCA already has issues with space.

If we come up with a new plane, then we will have to get a new processor. F35 used a newer processor than F22. I believe F35 uses 180/130/90nm processor. We can't use same processor for newer technology as it will require new architecture

You are exaggerating the need for power of processor. A laptop processor, for example use 30W of power with cooling for 2GHz dual core 22nm node. If it was 180nm, it would need 90W of power for same processing power. The 60W extra power is peanuts compared to the kilowatts worth of external heat. So, the internal heat will be completely eclipsed by external one.

Reliance got the largest share. That is why people were complaining. Not because the money went away. 50% was reinvested in India. The rest goes to Dassault. Not too hard to understand.
The most was given to DRAL. Smaller amounts were given to other companies like BTSL, DEFSYS, Kinetic, Mahindra, Maini, SAMTEL, etc. https://www.dassault-aviation.com/e...tract-india-clarifications-dassault-aviation/

Reliance got just 1000 crore. It is not the biggest share but just the "biggest known" share. We don't even know where the remaining 29000 crore went. The reinvested does not mean given in form of assets. It can simply mean that the money has been given back in one form or another. That is what I am saying as modifications to use UTTAM & Kaveri in Rafale which came from this offsets. Since it is under secrecy clause, it will never be revealed but this is most likely scenario

 

[gryphus-scarface]

Then how India upgraded into 180nm? Intel has enough fab facilities. In fact, intel does not make chips for mobile phones. SO, the idea that intel could not make enough chips for PC is ridiculous. If intel could make 100 chips last year and this year, the sales are down to 85 means 15% capacity is excess, not reduced. Since intel does not make mobile chips, there is no excuse for not being able to make chips for PCs

https://arstechnica.com/gadgets/201...t-as-pc-market-falls-to-58-5-million-systems/

Both Gartner and IDC say that there's continued influence from the shortage of Intel processors, caused by the company's long-delayed transition to 10nm manufacturing. That situation leaves Intel's 14nm manufacturing facilities overburdened. Gartner analysts said that these concerns disrupted the growth seen in the second quarter last year, as the delays prompted Intel to focus on higher margin products, with PC vendors following suit. IDC similarly cited the shortage of Intel chips at the low end as partly to blame for the market decline. To the extent that low-end chips were available, the PC companies seem to be favoring putting them in Chromebooks rather than Windows machines.

That is from the article. Intel could not meet the needs. They could not upgrade fast enough despite having the capital and being the world leaders in chip fabrication. Setting up a fab is not easy. Maintaining and upgrading one is even harder.

By the time the Rafale indigenisation sets in (after 36 rafales delivered in 2023), India will be having the capability needed

If you say so.

Stressful environment means:
1) External temperature going to 200 celsius. So, a chip must be made to endure 300 celsius temperature as a safety measure

Show me one chip that is designed to opperate at these temperatures.

2) The power supply is not some stable powerplant but the engine exhaust. So, minor fluctuations are expected

Zener diodes have been used as voltage regulators for a very long time now. They are taught in 12th. Voltage regulation and power backup systems are by no means new stuff or complicated.

3) Potential surge current when the plane gets hit by some missile or a bird

Why would there be a surge? What does a bird to to induce a surge? Also there's this really cool technology called surge protector. Look it up. Nearly every house has it.

In fact, shorting between nodes is one of the most common thing. Most processors have redundancy built in due to which the processor diverts the processing to some other node when this happens. All processors have some defective nodes but the redundancy takes care of that. As the nodes get smaller, the chances of defects increase. This is further exacerbated if external factors like temperature rises.

Shortage between nodes? CPUs having different nodes? What are you on about? A CPU is made on one node. The node size used to refer to the size of the transistor, but that relation has long gone.

Shorting happens due to:
1) defects in fabrication which in turn is higher in smaller nodes as the precision required is high
2) usage causing the nodes to wear out
3) temperature fluctuation causing the conductivity to rise and hence the leakage current to saturate

Defects in fabrication are eliminated by quality control checks. You can test the chip to see if it meets your required quality control measure. Usage? Usage rarely causes node wear out. How frequently have you seen commercial systems fail due to wear and tear of transistors? I'll confidently say that is never an issue.
Temperature fluctuations don't happen enough. Even when they do there is no evidence that it causes any of the isses you suggested. Anandtech has verified that 2012 Retina MacBooks could reach surface temperatures of 99C during operations. That was on 32nm node. Not military grade equipment. Let that sink in. A commercial air cooled system can maintain that without issues to its battery, CPU or any other component. Node size is not an issue.

The voltage being high enough is depends on what is "enough". When temperature rises, due to conductivity rise, the voltage requirement will reduce. So, what voltage worked in room temperature may not work. Also, the holes and electrons can move around faster when temperature increases which adds to degradation. All this is pronounced when the margin of error is very thin due to thin nodes.

Again these chips are liquid cooled, so their temperature will never exceed room temperature, let alone any temperature like 200 degrees. Degradation is not an issue. Stop saying BS. Give me actual sources for any of these claims. Yes conductivity increases with temperature rise, but the temperature can be controlled.

If we come up with a new plane, then we will have to get a new processor. F35 used a newer processor than F22. I believe F35 uses 180/130/90nm processor. We can't use same processor for newer technology as it will require new architecture

The F-22 already uses the PowerPC G5, that is somewhere between 130nm and 90nm. So the F-35 is without a doubt newer.

You are exaggerating the need for power of processor. A laptop processor, for example use 30W of power with cooling for 2GHz dual core 22nm node. If it was 180nm, it would need 90W of power for same processing power. The 60W extra power is peanuts compared to the kilowatts worth of external heat. So, the internal heat will be completely eclipsed by external one.

What is your point here? That cooling is no longer necessary? The lesser the heat output, the lesser. As for your claim that processing power isn't necessary, DAS already demands a whole lot of compute resources. Its practically AR systems, and with the compute power of early 200s systems, no wonder they faced lag issues. Let alon memory related isses. The best I could find is that it used a PowerPC CPU with altivec instructions, which already puts it in the same league as FOC F-22, and we know it improved from there. So there goes your theory that military jets used outdated CPUs, or that they somehow need older nodes. The F-22 in FOC used a CPU that was as new as possible.

Reliance got just 1000 crore. It is not the biggest share but just the "biggest known" share. We don't even know where the remaining 29000 crore went. The reinvested does not mean given in form of assets. It can simply mean that the money has been given back in one form or another. That is what I am saying as modifications to use UTTAM & Kaveri in Rafale which came from this offsets. Since it is under secrecy clause, it will never be revealed but this is most likely scenario

Did you even bother to read the articles? I'm tired of arguing with you. You provide no sources for any of your claims. Provide at least a few sources. For some of your claims. I can understand why you won't find any sources for Uttam on Rafale. At least for the stuff about the CPUs facing issues on smaller nodes.

 

[Advaidhya Tiwari]

Show me one chip that is designed to opperate at these temperatures.

Military grade semiconductors have temperature tolerance from -60 celsius to 125 celsius. The semiconductors will not freeze even in worst case scenario and will not get overheated as long as there is reasonable cooling. Don't expect 22nm to have 125 celsius tolerance

Zener diodes have been used as voltage regulators for a very long time now. They are taught in 12th. Voltage regulation and power backup systems are by no means new stuff or complicated.

Yeah and these diodes also are affected by temperature fluctuations. Everything is affected by temperature. When te temperature reaches 200 celsius, using cooling will be very difficult as that much power will not be available to get the cooling

Shortage between nodes? CPUs having different nodes? What are you on about? A CPU is made on one node. The node size used to refer to the size of the transistor, but that relation has long gone.

Shortage between transistors! The 22nm, 180nm is distance between transistors and they can short

Defects in fabrication are eliminated by quality control checks. You can test the chip to see if it meets your required quality control measure. Usage? Usage rarely causes node wear out. How frequently have you seen commercial systems fail due to wear and tear of transistors? I'll confidently say that is never an issue.
Temperature fluctuations don't happen enough. Even when they do there is no evidence that it causes any of the isses you suggested. Anandtech has verified that 2012 Retina MacBooks could reach surface temperatures of 99C during operations. That was on 32nm node. Not military grade equipment. Let that sink in. A commercial air cooled system can maintain that without issues to its battery, CPU or any other component. Node size is not an issue.

True, manufacturing defects can be resolved by testing. In case of smaller nodes, testing will eliminate large chunk of chips as defective if rigorous testing is made but that still can be resolved.

But wear and tear is quite normal. For example, you lose your pen drive, RAM or even processor in your computer (if you use it for more than 3 years) due to wear and tear of the transistors. As electricity flows through then, heat is geenrated and the transistors are affected. You must read about "degradation" of transistors

Again these chips are liquid cooled, so their temperature will never exceed room temperature, let alone any temperature like 200 degrees. Degradation is not an issue. Stop saying BS. Give me actual sources for any of these claims. Yes conductivity increases with temperature rise, but the temperature can be controlled.

No, even the liquid will need heat sink to throw out temperature. So, we can't expect it to be in room temperature. This is guaranteed. As I mentioned above, temperature tolerance of military grade semiconductor is 125 celsius. This is Wikipedia figure.

https://en.wikipedia.org/wiki/Operating_temperature
https://www.maximintegrated.com/en/markets/military-aerospace.html

The F-22 already uses the PowerPC G5, that is somewhere between 130nm and 90nm. So the F-35 is without a doubt newer.

No, PowerPC is an architecture and was used even in 1990s chips. Since F22 first flew in 1996, it can't be 180nm or below. I don't think F22 has seen any upgrade at all.

What is your point here? That cooling is no longer necessary? The lesser the heat output, the lesser. As for your claim that processing power isn't necessary, DAS already demands a whole lot of compute resources. Its practically AR systems, and with the compute power of early 200s systems, no wonder they faced lag issues. Let alon memory related isses. The best I could find is that it used a PowerPC CPU with altivec instructions, which already puts it in the same league as FOC F-22, and we know it improved from there. So there goes your theory that military jets used outdated CPUs, or that they somehow need older nodes. The F-22 in FOC used a CPU that was as new as possible.

Computer resource is needed for computing but there are multiple processors to handle each main equipment. It is not one processor for all. As I said, you can't get 125 celsius tolerance for 22nm chips. So, 90nm or above is needed.

Did you even bother to read the articles? I'm tired of arguing with you. You provide no sources for any of your claims. Provide at least a few sources. For some of your claims. I can understand why you won't find any sources for Uttam on Rafale. At least for the stuff about the CPUs facing issues on smaller nodes.

Ok. Source for:

DRAL getting only 1000 crore offset:

Reliance Defence will get 3% of Rs 30,000 crore Rafale offset

Reliability & breaking down of semiconductors:
https://en.wikipedia.org/wiki/Failure_of_electronic_components

About reliability of lower size nodes, below 90nm, the SiO2 is replaced with another high dielectric material like HfO2, GeO2. This gate oxide material change also reduces some amount of quality in terms of reliability.

 

[gryphus-scarface]

Military grade semiconductors have temperature tolerance from -60 celsius to 125 celsius. The semiconductors will not freeze even in worst case scenario and will not get overheated as long as there is reasonable cooling. Don't expect 22nm to have 125 celsius tolerance

Source that 22nm can't handle this? Don't claim stuff cite sources.

Yeah and these diodes also are affected by temperature fluctuations. Everything is affected by temperature. When te temperature reaches 200 celsius, using cooling will be very difficult as that much power will not be available to get the cooling

There is this fancy technology called liquid cooling. It keeps things within an acceptable temperature range.
Also, here is proof that a 32nm chip can work just fine at 75C : https://www.anandtech.com/show/6023/the-nextgen-macbook-pro-with-retina-display-review/12
That is consumer grade. And air cooled. Those systems are rated for 95-100C. Not a stretch to believe that its possible to achieve 125C for military needs.

Shortage between transistors! The 22nm, 180nm is distance between transistors and they can short

The question is why has this never been observed anywhere? I've never had any such issues with any of the equipment I've owned, nor have I heard of it anywhere. Your won sources don't mention it.

True, manufacturing defects can be resolved by testing. In case of smaller nodes, testing will eliminate large chunk of chips as defective if rigorous testing is made but that still can be resolved.

But wear and tear is quite normal. For example, you lose your pen drive, RAM or even processor in your computer (if you use it for more than 3 years) due to wear and tear of the transistors. As electricity flows through then, heat is geenrated and the transistors are affected. You must read about "degradation" of transistors

I am writing this from a laptop that was made in 2013. I've used it for at least 2 hours a day on average, with actual usage well above that. I've pushed its CPU and GPU to the limits with games and encoding various movies. Its CPU and GPU were made on the 22nm node. By my calculations, its roughly 6 years old.

I also own a PSP from 2006 that works just fine. A desktop from 2009 that's still functional. A laptop from 2002 that works just fine with linux. A Gameboy. I can go about systems that I have which are older than 3 years. The lab in my school had computers that were at least 5 years old. They all were used for 5hrs a day 5 days a week for 9 months of the year. They were replaced because they became obsolete. Not because they failed.

Yes transistor degradation is a thing. But it is irrelevant because of how long it takes. You seem to think that it affects a lot of systems. It doesn't affect anything. You will see mechanical parts like cooling fans fail long before the transistors fail. If you don't you ended up with a a bad chip. Nothing to do with node.

No, even the liquid will need heat sink to throw out temperature. So, we can't expect it to be in room temperature. This is guaranteed. As I mentioned above, temperature tolerance of military grade semiconductor is 125 celsius. This is Wikipedia figure.

https://en.wikipedia.org/wiki/Operating_temperature
https://www.maximintegrated.com/en/markets/military-aerospace.html

Huh, so you think that all those liquid cooled set ups are fake? You do realise that liquid cooling is even done on some high end gaming rigs? And your Wikipedia figure says 125C is the broadly accepted value. Lets take that as the value needed. Show me evidence that a 22nm node can't reach that temperature and work. I've shown evidence that commercial CPUs can operate at 75C just fine. The emergency shutdown for commercial Intel CPUs is 95-100C. So that means they will work even at 90C, just very close to their limit.

No, PowerPC is an architecture and was used even in 1990s chips. Since F22 first flew in 1996, it can't be 180nm or below. I don't think F22 has seen any upgrade at all.

I specifically mentioned the PowerPC G5.
Here is the Raytheon press release :

http://investor.raytheon.com/phoenix.zhtml?c=84193&p=irol-newsArticle&ID=439701

The original lots (I think up to lot 5) used the Intel i960MX which was deemed outdated. Later batches used the Raytheon CIP. This thing uses the PowerPC G5, aka PowerPC 970

Another source for this :
https://www.militaryaerospace.com/a...electronics-but-plan-for-future-upgrades.html

Here is a source for the PowerPC G5 / 970 : https://en.wikipedia.org/wiki/PowerPC_970

Computer resource is needed for computing but there are multiple processors to handle each main equipment. It is not one processor for all. As I said, you can't get 125 celsius tolerance for 22nm chips. So, 90nm or above is needed.

Yes not one processor for all. However doing something like DAS is very compute heavy. Why do you think we didn;t see AR/VR systems until 2015? Even then they demanded the very best CPU + GPU combo for a good amount of time. DAS is in some ways an AR system. It takes input from its sensors and then paints this world onto a helmet. It needs to track the user's helmet movements and show them the respective area they want. It also needs to overlay other information onto this, like target details, on what it has identified. This is very much compute intensive.

Again give source that 22nm physically can't work at 125C, but other higher nodes can.

Ok. Source for:

DRAL getting only 1000 crore offset:

Reliance Defence will get 3% of Rs 30,000 crore Rafale offset

Yes DRAL got only 1000 Cr. I am not disputing this. I am disputing that the rest will go into Uttam for Rafale. see my links. The Hindu actaully has a time table on when MBDA, Thales, Dassault will release parts of their offset, and to which company. The first year offsets of MBDA and Thales are 0. They pick up later.

Reliability & breaking down of semiconductors:
https://en.wikipedia.org/wiki/Failure_of_electronic_components

About reliability of lower size nodes, below 90nm, the SiO2 is replaced with another high dielectric material like HfO2, GeO2. This gate oxide material change also reduces some amount of quality in terms of reliability.

Again you seem to think that lower node size = less reliable. My laptop is just fine today. And nobody to date has complained because their CPU short circuited itself, or its transistors degraded.

And just as an FYI, F-22s have gotten a lot of upgrades, just nothing that is major, or has redefined their roles.

https://defensesystems.com/articles/2017/03/14/f22.aspx