PAK-FA enters Korean Fighter competition, 4 real

http://www.flightglobal.com/blogs/the-dewl...a-to-the-o.html

So that makes it a race between the Eurofighter, F-35, Silent Eagle and PAK-FA.

THIS SHOULD BE INTERESTING.

Now if China would only enter the J-20

Surely it is interesting. Maybe after this we will get some more news about the PAK-FA's actual peformance compared to others.

My guess is that SK will go for the SilentEagle though. Less expensive, more suited for a small country like South Korea.

ROK already uses a small number of BMP-3s and late-model T-80s. It's nothing new, and Russia is probably giving them the PAK FA to give them another avenue to pay off their debt by.

Those traitors. Those pathetic traitors.
Comrade Lenin is turning in his grave right now. And considering that this grave is an open glass box its only a matter of time before it becomes a youtube meme.

"Welcome to the new Russia, captain Purp."

JK.

Nobody could consider an F-15 less expensive, the running costs of those things alone should put it out of contention.

Nobody could consider an F-15 less expensive, the running costs of those things alone should put it out of contention.

Except S. Korea already has a well established F-15 support infrastructure and to be fair none of the competing options are really markeed as cheap to operate, well appart from some of the more optimistic F-35 stuff.

IIRC the Koreans have been playing hard ball with the russians in regards to debt repayments and have been refusing to accept arms as payment (the total shambles that is Metis-M might be part of it).

Buying F-15 seems short-sighted to me. Far East is going to be an 'active' region in the service life of the plane, so they should buy a good one rather than a cheap one. PAK-FA actually looks the most appealing unless its claimed capabilities are outright lies.

In modern fighters (so-called 5-gen, although I'd say F-22 and PAK-FA are at most 4.5-gen), avionics comprise an overwhelming proportion of the cost. The airframe, even a stealth one, is not all that costly. Considering that, and a $100-million cost estimate for F-15SE, it's likely that the cost difference will not be a major factor.

The problems with PAK-FA are largely the same as the problems with AMD Bulldozer: "WHEN will it finally come out?" and "What exactly will it be?" - public information is, compared to Western developments, quite scarce.

More specifically, PAK-FA is being designed around a very modern and sophisticated avionics suite. There are already CPU and system architectures for it, with more performance than even the avionics slated for F-35. The problem is with actually making these chips. They require a modern high-complexity semiconductor fab, which very few countries have.
Since mistakes have been made in setting up the appropriate fab in Russia, it's not certain if it will be able to deliver domestically produced chips in time.

Currently the chips are being made in Taiwan, which presents a political problem, since neither Russia nor South Korea recognize Taiwan. It's one thing when a commercial company orders chips from Taiwan, but a whole other for governments, and it's not trivial, up to as much as 3-5% of value might have to come from Taiwan. Other than Taiwan, all other pure-play foundries are located in or controlled by companies from either China or NATO, which can cause even more political problems.

A full-on partnership with Korea could potentially save PAK-FA's schedule, since Hynix and Samsung are much more capable of setting up adequate manufacturing facilities (they are not, however, currently ready to produce the chips on order). But such a partnership has deeper political implications for both countries than just a purchase, so appropriate politics would have to be worked out first.

I thought that they were putting MCST chips on it?

Would Taiwan's involvement in the supply chain be that big a deal? S.Korea already trades quite a bit to Taiwan, and the PRC would look really awkward complaining when they are still dependent on Russian jet engines for their domestic and even export aircraft.

They are MCST designs (for primary CCM, there are other chips as well, from other companies). But the thing is, they aren't put in silicon by MCST, Angstrom or Micron.

There is a 90nm fab that was supposed to be transferred to Angstrom a while ago, but the actual transfer wasn't done properly, since they tried to refurbish an old cleanroom building instead of building a new one on empty field. Currently it's supposedly sort of installed, but isn't producing anything yet. It might be that it never will, these things are possibly the most finicky and fragile creations ever produced by mankind.

Currently the chips are produced by TSMC. PAK-FA is supposed to use 65nm chips if possible, current ones are 90nm, and a fallback plan to 90nm is available.
Meeting at least the 90nm objective is critical for completion, since T-50 has a very limited avionics bays capacity for its timeframe. Current airframe's avionics bays measure 1.5m^3 for the radar (might not be met), a bit over 1m^3 for analog systems, 1.7m^2 for auxiliary systems, 1m^3 for primary cooling systems, and 2.5m^3 for primary processors.

PAK-FA avionics are supposed to be packed tighter than usual, likely using sandwich arrangements and/or through-PCB cooling, but even at increased density that's still only 2,200-2,800kg, compared to over 3,000kg already exceeded in some modern fighters. The software required to run its systems depends on latest and developmental processors, it won't work if the hardware is only halfway there.

And it's not clear how the required chips are to be made, because ordering them from a country that supposedly doesn't even exist is politically difficult. What is produced now is not really mil-spec, it's commercial equipment in commercial packaging. Moving it to 65nm, HTCC packages and military QC wouldn't be a technical problem, but it would be a different product, and it would need to be ordered by governments. As such, government interference is possible from any of the three sides involved, undermining the likelihood of successful delivery.

Related: How much do you know about the state of MCST and other such companies?

A few bits here and there, nothing in detail really. Regarding anything in particular?

Specifically, is there any expectation of any of them producing consumer-grade computer components any time soon, especially RAM and CPUs?