In a move that both validates the technology underlying Voice over Internet Protocol (VoIP) and promises to help advance it (especially in terms of security) the U.S. Department of Defense (DoD) has taken a step toward developing its own, privately managed VoIP system. Nortel Networks recently received a contract worth up to US$20 million to migrate the department's network to one that, through installation of new software and other upgrades, will be able to support VoIP. More details at Defense Industry Daily.
I've also noticed that the DoD is investing significant research in Gallium Nitride (GaN) semiconductors. GaN is a next-generation semiconductor material that would produce smaller, lighter, more efficient and more cost effective systems than can be realized with current technology. Its 10x output power gains offer the higher power density and efficiency required for high power phased array radar, electronic warfare, missile seeker and communications systems - and eventually, advanced consumer electronics.
American tax dollars at work for you...
the problem that I have with all of these hi-tech systems is that the military may wind up becoming dependent upon them -- and that there may be relatively "low tech" ways to interfere with or disable these systems.
Iraq has amply demonstrated the inadequecy of dependence upon high-tech solutions to address the real issues faced by troops on the ground. The problem is not our lack of ability to destroy things and kill people, but what happens after we've destroyed what we need to destroy and killed who we need to kill --- and it does not appear that the answer is to destroy more stuff and kill more people.
Iraq was not a military problem, but a political one, and its becoming increasingly obvious that our problems in Iraq are the result of trying to address a complex political conumdrum with the military. We can build all the weapons systems in the world and still not be safe as long as the political decisions are being made by people who are unconcerned with the limits of military force to address political issues.
Iraq was both a military and a political problem... and research gives the u.s. military greater capabilities and hence greater options in future situatios.
Issues of security are of course highly secondary to basic research such as Gallium Arsenide chips - that's something you design into a chip or system once the technology of making such chips works.
Security is exactly the plus in the VoIP case - as the military will be advancing the state of the art. This will improve corporate adoption of VoIP, hence available technologies, hence lead to more mature solutions, faster, that can reduce your phone bill. While saving the military money and moving it another step toward packetized communications that are - wait for it - more difficult to disrupt.
So I'm wondering what p.'s problem is, exactly - other than the existence of the u.s. military itself.
Back when I wasn't paying very much attention to politics, a member of the Green Party, and an avid Ralph Nader voter; I was all up in arms about the money sink that was the industrial military complex.
Then I saw the fruit of that military development in Afghanistan and Iraq. How far we have come from the days of carpet bombing and napalm. War is certainly hell, but it is hell for far fewer civilians than it used to be thanks to the industrial military complex.
Given the reality that we do not live in a can't we all just get along world, and that military conflict is inevitable, I think military research and development is essential, even humanitarian. I also believe nations that are still riding on arsenals of old dumb bombs and napalm (most of Europe, Canada) are shirking their humanitarian duties. I also don't begrudge the military spending as the U.S. is not even the biggest spender, in terms of percentage of GDP.
Joe
Perhaps it could be innovations such as this. It has been released upon the world and as far as I can tell has had deep and profound affects on personal lives concerning involvement with their government as well as the inhabitants around the globe.
Any comment on this?.
What would be ways to reform this process?
Re: "Nortel Networks recently received a contract worth up to US$20 million ..."
Which Nortel promptly entered into the books as $200 million.
I guess the real question is how much are you willing to have your taxes raised to support these kinds of programs.
Nortel Networks VoIP government contract.
Federal government WIC grants.
Frequently asked questions #10.
10. How much does WIC cost?
Congress appropriated $4.696 billion for WIC in FY 2003. The appropriation includes $25 million for the WIC Farmers' Market Nutrition Program.
By comparison, the WIC program cost $10.4 million in 1974; $727.7 million in 1980; $1.5 billion in 1985; and $2.1 billion in 1990.
Care to ask again how I want to cut the budget?
FWIW, I will just point out that the two programs Joe has bracketed together are quite different. The Nortel/VOIP deal represents the .mil guys adopting a technology that is already being rolled out in the commercial world. It's almost - but not quite - a COTS (commercial off the shelf) program, which is considered quite the thing to avoid paying for too much one-off, milspec gold plating. The points of difference are likely related to security necessities.
GaN, on the other hand, is much more of a basic R&D program, necessary because it's one of those semiconductor materials that always looks like it might be a coming thing for commercial products, but never quite gets there. ('There' being a self-sustaining reinvestment cycle with scale economies.) Cree is a company that seems to specialize in that kind of semi-orphan materials, which are often capable of tricks for which you're willing to pay when lives are at stake, but willing to compromise when it's a question of a commercial product's BOM cost.
Antics on both sides of the house and things like this don't help either. Military Bill Carries Range of Extra Spending
BTW folks thanks for sponsoring the Washington Nationals.
Look for diamond based semiconductors as well.
The process was developed to coat the surface of hard drives.
Instaed of a painted on mixture for the mag surface our hard drives are a layer of metal made by boiling metal in a vacuum, where the energectic atoms (due to their heat) smash into the aluminum surface creating a very uniform coat that is only atoms thick, over this they layer on a coat of diamond.
Vac deposition of diamond has already been used to create really large gemstones. and natually you can also grow wafers of the material.
Diamond dont have the heat tolerance problems that silicon substrate does, and its a far better conductor of heat.
Changing nothing else, you can make a cpu that can be clocked 20 times faster.
It also has far less of an electromigration problem, glass is like a taffy that never really becomes a solid, in those 1000 year old churches, look at the windows, and you can see the glass has mesurably flowed.
This is why cooling your cpu an extra 20 degrees gives you triple the lifespan.
A diamond chip would not have this problem, it would mean a space probe sent out on a 10,000 year mission could still be working, whereas with silicon the chips have become non functioning specks of colored glass.
If some of you ever wondered why you have owned gadjects that went bad one day for no reason at all. its at least half the reason for those failures (the capacitors drying out is the other)
The next thing we need to do is more EMP hardening of our stuff. AMD already has the tech in their hypertransport serial bus tech for a complete move to optical interconnects.
Care to ask again how I want to cut the budget?
no, the fact that you prefer to starve children so that you can have better reception while video conferencing says it all.
That's the kind of ridiculous rhetoric that makes you no better than Ann Coulter yourself, lukasiak. Your hypocrisy is noted.
SiC is a better near term bet for increasing the effectiveness of power electronics.
And it is already in production - blue and white LEDs.
The problem with diamond at this time is that the rate of surface defects per unit area is still too high for production use.
One of the things we need to do to improve CPU efficiency is to couple the language used for coding more tightly with CPU design. Fewer transistors toggling per cycle.
The "C" language is very poor in that respect.
A language like FORTH designed originally to run on a virtual processor is much better in that respect.
Also the smaller the processor the fewer defective ones per chip. If most of the chip area (for multi-million gate chips) is devoted to on chip memory, bad memory sections can be replaced with working sections (now done with large RAM) you can get better production rates with a high defect material.
If the code is mostly high use small sections speeds can be dramatically improved.
The penalty - memory management must be done by the software designer rather than by energy using, hard to design silicon.
A language like FORTH designed originally to run on a virtual processor is much better in that respect.
There you go again. <reagan/> 8)
FORTH running as a TIL (threaded interpretive language) makes poor use of registers, is not type safe, is not inherently parallel.
C (and C++) compilers, as well as java and C# JIT compilers, have made many advances over the last decades. There really isn't any comparison.
Note also that with the AMD64 architecture (adopted also now by Intel), there are more general registers to play with, making a compiler even more important.
Vint Cerf recently stated the biggest issues for computer science now are security and scalability, and I think he’s right.
That so many attacks happen due to a lack of buffer limit checking is a scandal. The NX bit (another AMD innovation) is a good step. Type safety and type checking in the language will help. But there is a long way to go.
With regard to scalability, this article by Herb Sutter will be viewed as seminal in a few years. Herb is right, the free lunch is over. In the future, you’re program will have to be parallel if you want it to go faster. (See also this article if you are a Scientific American subscriber). Again, language and compiler issues will be used to address this. Recall that parallel programming has been the wave of the future for thirty years. The future is now.