What I heard is that Russia failed to secure the order for the Su-30MK3, but there is another order of 24 Su-30MK2 for next year. The reasons for failure of procurement is not known but it may have something to do with the time and length of delivery. China wants it fast, not some years down the road. Slow and delayed deliveries have tainted Chinese impressions of buying from Russia—orders for the two Sovremannies and the 8 Kilos are fairly late. Russia appeared unsuccesful in securing orders for upgrading Chinese Su-27s to the Russian template Su-27SM standard, or to convince China to upgrade to the Su-27SM. The main reason for this is that Russia wants to sell the avionics and radar, or the Chinese do not want to pay the bill for licensing them any more, preferring to upgrade the Su-27s with their homegrown package.

I need or want any clarification if any new avionics equipment are bought by China.

I knew you’ll bring in Russia bashing somehow. 🙂

I gather you had secret access to Chinese avionics labs to know that all of a sudden China abandoned Russian based tech to move to Western tech. As plawolf himself said, China hasn’t seen Western stuff for a long time due to sanctions. There is every reason to believe that what was is continuing.

Quite incorrect. Everyone with a common sense knows that the Israelis have been crucial in advancing China’s techs, especially with military electronics. ELBIT even has an address in Beijing.

IOW, it is you who has to prove that China somehow has changed avionics and radar tech in the last few months.

China’s avionics have advanced sharply between 1995-1998. Before 1995, the techs appear 60-70’s. By 1998, planes were sporting sophisticated electronic HUDs, MFDs, and slotted planar array radars. The quantum leap could not have been done indigenously so fast but with outside help—the Israelis.

So yes, China’s avionics tech have advanced to mid Western standards (not high end yet), fairly enough to the early to mid APG-68 level or APG-66 MLU level, thanks in part for having such good teachers.

That advancement is quite keen since China never pursued contracts to buy Zhuks for the J-8II and J-10, Kopyo for the FC-1, Russian radar for their AWACS, or even the Russian recommended upgrades for the Su-27.

Alternative name for A-50 Golden Eagle is Wei Kuo, foster brother of Ching Kuo. That’s because 100 engineers from AIDC involved in the Ching Kuo went to work on the A-50.

Crobato

If you think No. 37 is LO, you are quite mistaken. See that big bulb there? That gleams radar off like a beacon. The first thing that is a no no in an LO design is 90 degree surfaces. The second thing is round and curved surfaces. No. 37 has the bulb, the spherical turret, and the cylindrical AshM launchers. If you cannot understand the disvalue spherical and curved surfaces are to RCS, don’t bother talking about “carefully contoured”.

Curved surfaces can produce varying effects on RCS dependent on the design – they may not reduce RCS but they can de-emphasize particular zones on a ship and magnify others to confuse, for example, a missile seeker as to ship aspect. The comment you made is overly simplistic.

I’m not refering to the curved surfaces on the body. The rounded, cylindrical and spherical surfaces you see in the bulb, the turret and the AshMs certainly do not help in reducing RCS, rather they worsen it.

I don’t see how the housings in 2208 being RCS increasing. Rather, their angles are quite correct, you need to throw the radar vertically to the sky. In addition it most probably has RAM.

Compare that to ship no. 37 there. That ship is a lot more unstealthy. There is a preponderance of spherically shaped surfaces like the radar bulb. In addition the turret shows rounded surfaces.

Anyone who does not house their ASHMs launchers like No. 37 there increases their RCS as those cylinders are going to reflect radar major. Most of the other designs I’ve seen has no provisions for large AshMs and are more like design exercises. The moment you put the launchers, there goes the RCS.

Someone needs an education about the architecture and reliability of the F-22’s CIPs.

Critical functions, such as integrated flight & propulsion control, cockpit display management, and redundant subsystem management are split between the 2 CIPs. Even if one is totally destroyed, the F-22 can fly home and land safely with the remaining CIP. Non-flight critical functions such as sensor management, EW, and ICNIA can be hosted on a single CIP.

Each CIP carries spares for each type of module and each module has comprehensive continuous BIT. The use of these spare modules is controlled by a failure management function which plays traffic cop and assigns functions to each module in accordance with predetermined levels of importance. Should a module fail, it’s Operational Fight Program (OFP) is downloaded into one of the spares from a memory module. The spare takes over the functions of the failed module. An F-22 can continue to generate sorties until the last spare module is tasked to replace a failed module. Should multiple modules fail during flight, the failure management function sheds the least important function and downloads the higher priority OFP into that module. The failure management function also provides an in-flight module reboot and software reload function that can be used to restore module function.

This fault tolerant architecture gives the CIP a Mean Time Between Loss of Function (MTBLOF) an order of magnitude better than anything flying and allows CIP maintenance to be deferred to a convenient time (when the pace of the air campaign slacks off).

And you need a basic education about computer architectures.

The system you are describing above is a highly modular one, where different modules handle separate tasks, coordinated through a databus by a controller, which is the CIP. This is not a monolithic design where a single CPU—processor—is handling all the subsystems with a single memory body (classic Von Neumann architecture). It is also not unknown (rather typical in fact) of large control systems like in airlines, tankers, and manufacturing robots.

The CIP is not a processor per se like an Athlon or Pentium 4, which is what you properly call as a CPU or Central Processing Unit. Do note that the military lingo do not jive with most tech talk, since sometimes, a complete system component with CPU, memory, and I/O is called a “processor”, when most people understand it as just being the CPU itself. It is a complete computing system by itself, interacting with other complete systems through a bus, more like a distributed network, rather than a monolithic layout.

Hardware redundancy does not protect you from software failure. The same bug can occur in the same mirror backup module. Software failures are simultaneousl—if the same CPU handles the same set of subsystems, a failure—usually a bug—-in one subsystem or in the software code takes the entire system.

As for now, frankly, the CIP has not achieved what you are claiming it to do, since it only has a MTBF of less than 10hrs, and previous to that, was struggling with about 1 to 2hrs.

If you guys think the NSA or CIA are willing to spend $50,000 for “PER” IC in low volumn and get just commerical grade custom circuits then, there is a lot about secret programs the US Military/Government you need to learn.

That should be right for a low volume chip that is only made in the hundreds, not hundreds of thousands. If you go to a freeplay fab in Taiwan and ask for something in that quantity you will get bill. But it does not mean anything like 10GHz. You can make a 10GHz chip cost like a thousand bucks if you make enough of it, and 10MHz chip in the tens of thousands if you only produce so much of it.

But that not for a chip that you would have to create a special fab or process to make. In which case the total number of chips made would have to be equal to the cost of the fab, the R.D process and the corporate overhead. We’re talking of tens if not hundreds of millions here.

By the way, obsolete chips do cost a lot to make nowadays, because you have to keep an obsolete plant with a 0.8 to 0.5 micron process just to make them, and without any other other viable commercial product to support such a plant, those chips have alone to shoulder the cost and overhead of keeping such a plant alive.

Do you think Intel, AMD and, other companies pay all their R&D cost they need to educate themselves. (That will be difficult because this sort of stuff does not show up on company statement for investors to see. These are projects which use special development diffusion furnaces.

Crapola. You do need to realize that bulk of revenues Intel, AMD and etc,. makes are in the commercial sector. Try reading their financial statements next time. These statements are prepared by third party accounting firms and would have very much to answer for. So yes, you cannot hide “unseen” sources of revenue from public eyes.

The technology goes where the money is.

http://crpit.com/confpapers/CRPITV6Beckett.pdf Page #3
EXCERPT: [These are technicle papers from the previous decade.]
Finally, semiconductor behaviour has recently been demonstrated within very narrow carbon nanotube (fullerenen) based structures (Wilson et al 2000). Nanotube technology may eventually supposrt the construction of non-volatile RAM and logic function at integrated levels approaching 10¹² elements/cm², operating frequencie in excess of 100GHz

Nanotubes are still not applicable processes. This does not prove your assertion that the F-22 has 10GHz CPUs by one bit.

When North Korea captured the USS Pueblo, do you think the Soviets made such a lucrative deal with the North Korean for circuit they could have purchased through some “dummy company?”
When the P-3 Orion landed at the PRC air base, do you think so much effort to remove all avionics on the P-3 if they could tell how advanced American semiconductors were through commercial means?
Guess what the CIA and NSA are purchasing in the way of technology now? BE REAL!!

You are the one who needs to get real. They’re not taking apart how to see the semiconductors work, but how the entire systems work, especially with the encryption.

You can see how the US government is so behind in its idea of tech that it considers chips inside any Power Mac G3 and Sony Playstation 2 to be nonexportable to China.

Here in the Valley reverse engineering is very common.

Not without a major lawsuit.

The F-22 is a different animal, its radar changes frequency and power level with each pulse and the F-22 has only one computer, not a computer for each major systems. The F-15 has a flight dynamics, fire control, radar and, a couple of others. The F-22 has just the two computers that the USAF says has the computing power of a cray supercomputer, that statememt is made for a reason.

Show me that statement man. That’s really a bad design you’re talking about. For real time control processes, which an aircraft is, the best architectures are modular highly distributed systems. That’s one reason why you have buses like the old Mil Std. 1533 spec. With monolithic systems, you have a high chance that a failure will take down EVERYTHING. There is no redudancy against failure or battle damage.

By the way, many commercial chips now have the power of a Cray supercomputer of the past. Why do you think they attempted to put Power Macs in the no export list to China?

Use sand paper and leave marks, if you are trying to fool someone and that someone have access to powerful microscopes. People who take this sort of approach when they try and run a skeem on a customer and get caught! Being a design draftsman when these skeems go wrong, it affects management and marketing. It is so funny to watch the “fit hit the shan!”

Sand it well enough, and you can take the labels off. Commercial chips are so dimed down nowadays, companies don’t really want to spend a lot of money on label dyes so it will stay there permanently.

> You’re just preaching, that’s all I can say.
Yes but from someone who has worked in the industry for over three decades and has firsthand knowledge of which I speak! If you believe commercial circuits are ahead of military technology then you will probably believe the latest Air Bus or B-777 have later technology than that the military use.

And yes, you can believe that the latest Air Bus and B-777 have later techs than the military, hence why some of the latest Boeing planes have been export banned from China because a certain chip there used in the navigation systems are the same as those used in cruise missiles. Maybe you need to figure out why some of those commercial planes cost as much as, or more than an F-22. Do you think GE’s most powerful turbofan engine is used in the military? Guess again.

The issue about chip export restrictions has angered many developers as this has caused their products to be cut out from many markets. This may even prompt a move to create chips that they won’t sell to the US military, so there will be no export restrictions on them.

> aerospacetech
> my information comes from a Raytheon publication. You might not
> like it, but the truth is the F/A-22 does use old processors.
“Let us agree to disagree!” There is conflicting information on the Internet about it.
There is some information saying it is a 25MHz computer and other information saying it is a custom chip from Silicon Graphics!

Silicon Graphics owns the MIPS RISC architectural design. MIPS are common chips used in microcontrolling applications. You can find them in Playstation 1 and 2 among others. That Nintendo box before the Game Cube had a MIPS chip and even used the SGI connection with their ads (the chip is a 100MHz R4000 with a 64 bit bus). One of the most common MIPS processors for embedded use is the R3000 which has a 25MHz version. Yes, variants from this family is used in the Playstation 1 and in the modular mission computers in the F-16 MLU.