Brumby
Major
The point is not about speculation per se but which is a better explanation. The more significant the claim the greater the corresponding demand on it to demonstrate its claim rather than not.
Chinese Radar Developments - KLJ series and others
- Thread starter indochina
- Start date
The point is not about speculation per se but which is a better explanation. The more significant the claim the greater the corresponding demand on it to demonstrate its claim rather than not.
I feel like the original blog post has made TR module density a massive issue, as if it is some kind of valid benchmark to compare between two different industries.
The only really fair way to judge would be to make two countries start development of a new radar simultaneously and have a pre set end date for them to present their products. But in real life what tends to happen is that one guy A does things first and produces X product, and the other guy B a few years later can leverage some newer technology and their own R&D to produce Y product which is very similar to X and possibly a little better depending on how far behind guy B was and how much they invested into the type of technology compared to guy A, as well as the nature of the technology itself.
I'm simply working off my understanding of technological development. There's a presumption about technological development as a linear process that's actually quite unfounded if you study the history of technology.
I wouldn't put it that way. Keep in mind that the comparisons used in the article are not new radars by any means. It's unlikely that the US isn't already working on a radar with greater module density. If there is a miniaturization limit though, I suspect it will have to do with cooling limitations and not module density.
But what makes that argumentative demand an imperative is not standard of fact but believability. We're dealing with a particularly opaque subject that will perpetually frustrate those standards of believability. That does not make the claim less likely to be true.
Last edited:
Logically I agree with you. Practically, and especially with regards to PLA watching, I think there are some maxims that must be bent a little, simply because of how little solid and reliable information there often is to "demonstrate" claims. I mean, if we went by that kind of logic, one could conceivably argue J-XX/J-20 was just simply the collective dream of PLA fanboys prior to us actually getting clear photos of it.
The way I deal with these kind of speculation is to make it clear that nothing with PLA development is ever set in concrete. We are always getting new rumours, new news reports, new photos, which may change the balance as to how likely scenario A is, or makes us assess or re assess a premise which might be necessary for a scenario to occur.
Last edited:
Well, the adventage with electronics is that you can skip old technolgy and moves to the state of the art technolgy. No one would need to "master" the art of relays before producing computer transistors in 45nm for modern computer chips.
Unlike metallurgy for example.
And right now China is capable of producing AESA radars with comprable T/R module to size modules to the USA or Europe. That would be only possible if China is able to solve the same power and heat management problems.
Also no one would claim that the several European countries aren't capable of producing state of the art AESA radars for fighters although their AESA radars are still in developement (Eurofighter, Gripen) or in the phase of introduction (Rafale).
Unlike metallurgy for example.
And right now China is capable of producing AESA radars with comprable T/R module to size modules to the USA or Europe. That would be only possible if China is able to solve the same power and heat management problems.
Also no one would claim that the several European countries aren't capable of producing state of the art AESA radars for fighters although their AESA radars are still in developement (Eurofighter, Gripen) or in the phase of introduction (Rafale).
Ultra
Junior Member
I think you got it wrong. Packaging elements smaller and smaller is not going to be easier and easier - it is actually harder and harder as it requires heavy research investments and a very large talent pool - which are all very capital heavy. It is similar to CPU design where US has majority of world's CPU design talents (Intel, Marvel, Texas Instruments, Nvidia, Qualcomm... etc etc just to name a few). As CPU gets smaller and smaller, investment in research to discover new materials and processes in order to further miniturize the CPU becomes critical - for example, going from 32 nm down to 28 nm is a huge jump as silicon starts to leak voltage and CPU becomes unstable - this leads to the discovery of High-k Metal Gate Transistor which use a different material entirely, and requires enormous investment to discover the correct material. Going from 28 nm down to 22 nm requires further packing of CPU circuitry by going 3 dimensionally (Inte's Tri-Gate Transistor Technology). Going down to 16 nm or even 14 nm requires inventing new processes - like the double patterning liquid immersion lithography.
Right now, the best Chinese foundry can produce is 32 nm and 28 nm is just coming online - so China is couple of generations behind.
antiterror13
Brigadier
ohh well, most Xeon CPU for supercomputer are produced with 32nm technology
Ultra
Junior Member
Building a world beating super computer nowadays is more about money than anything else. China has a lot of it and it wants prestige so they bought a lot of high-end intel CPUs and hooks them all together and there is the Tianhe-2
The truth is, if China is actually capable of designing a world beating super computer of their own, they would do so with THEIR OWN CPU (and not rely on Intel which is american).
vesicles
Colonel
I think you misunderstood his point. He meant that the packaging becomes easier WHEN the critical elements become smaller. I don't think he meant to say that it is easier to make elements smaller.
Engineer
Major
The most difficult part is hooking them together. As the number of CPUs increases, the quality of the CPUs themselves become secondary. Obviously, putting together the world's fastest supercomputer isn't something everyone can do, or everyone would have supercomputers.Building a world beating super computer nowadays is more about money than anything else. China has a lot of it and it wants prestige so they bought a lot of high-end intel CPUs and hooks them all together and there is the Tianhe-2
China does have homegrown CPUs, but most software are written for American CPUs with X86 instruction sets. It would be pointless to build a supercomputer that has no software to run.