No, the USA doesn't have that concept (mature age student) in any formal way. You are treated no differently from a new high school graduate. However, most "Mature Students" would go the Junior College route, which makes it significantly easier to get into a given state-supported school. I agree that older students are probably more likely to be motivated, my own personal observations supports that. And as for dropping out at 16, then getting back to school, and being successful all the while... may I only say "GOOD ON YA!"
Uh.... what, exactly is tactcile feedback? Maybe you meant to say tactile? I think you meant to say something completely different, and in your effort to use big words said something completely asinine
He meant testicle feedback. It is a new gaming paradigm that involves a car battery and some electrodes or some-such...
He was also a very nice man. I knew him ten years ago when I worked at Lawrence Berkeley National Lab. The first time I met him was when he ran into me (literally) in the cafeteria and he said "excuse me son". This is amazing for a up-and-coming PhD student. Wow.
The DOE runs our system of national laboratories, and is the successor to the Atomic Energy Commission. They aren't all that concerned with gasoline, as that is a small part of their work. They mostly work on nuclear weapons, fusion research, high-energy physics, renewable resources, etc. I used to work at Lawrenece Berkeley National Lab designing subatomic particle detectors. I couldn't give a rats ass about how much you spend for gas.
A processor does do significant work. For example, the electrons in the channels of the transistors move against some finite resistance, phonons are released as electrons fall from the conduction band into the valence band, etc. It only doesn't produce light because there is a momentum change (hence the phonon) when the electrons do so. In direct bandgap semiconductors, light is released, hence LEDs and semiconductor lasers. The reason there is heat in any processor is the fact that unless there is some resistance somewhere, the clock rate could be infinite since it would imply that any capacitance could be charged up in zero time.
In ten years a few engineers will be designing new classes of electronics based on quantum principles.
As another replier pointed out, the silicon transistor is based on quantum principles. In particular, it's operation cannot really be understood without some knowledge of the Quantum Theory of Solids, which was originally proposed by Fermi, among others. Basically, only through statistical mechanics and other quantum stuff can you show that there are such things as "minority carriers" in semiconductors and that they exist in enough quantities to make a useful current. I think you meant that the theory of computing would be quantum, not the electronics with which it is realized. Also, lasers are based on photons. And we have memory devices in the lab based on spin (didn't you mean electronic spin rather than magnetic spin?)
So, if a guy wanted to build a more "modern" homebrew CPU, what options are there? Are there any decent CAD tools that don't cost a thousand million dollars? And once a layout is done, is there anywhere you can get just one single chip made for a reasonable price?
LesPaul,
I built a CPU using freeware tools as part of my PhD project. The paper is "A 12-bit 80MS/s Pipelined ADC with Bootstrapped Digital Calibration" published in the IEEE Journal of Solid-State Circuits, May 2005. You can google the title if you want to see the paper. Anyway, I designed a 24-bit microprogramed CPU in 0.25um CMOS to act as the calibration controller of the Analog-to-Digital Converter (ADC). The project was great because I got to design the architecutre of the CPU, the microcode/instruction set, I wrote a custom assembler, etc. I designed the circuits using viewdraw (public domain UNIX schematic capture tool). I designed the logic and tested the circuits using a public domain VHDL simulator (can't remember which one.. alliance maybe?). I laid out the circuit using Magic, a public domain layout editor running on Unix or Linux. The only thing that cost money was fabricating the chip. There is a service called MOSIS (www.mosis.org) that will do multi-project runs to lower the cost for you. I think the cheapest you will get is a couple of thousand bucks for 40 parts or so. Mine was something like $40k but I had high performance analog circuits in a fancy process. Email me if you need more info at carl.grace@yahoo.com
But see the problem is when the other companies have people "over a barrel" or are outsourcing, then they have an advantage over the one thats playing nicely with the union. And eventually the unionized company falls behind, more than likely because they can't provide as many incentives to investors so they can't grow like the non unionized companies.
That's what trade unions or industrial unions are for.
Of course (TANGENT ALERT!), the remaining reference to "natural phenomena" still precludes even the question "is computer science a science?"
EngrBohn: I've always thought that disciplines with the name "science" after them are never really a science in the strictest sense. They are more likely applied forms of another discipline, and while valuable, not a science in their own right. For example, Material Science is really applied solid-state physics, Political Science is applied sociology, and Computer Science is a mixture of applied mathematics and engineering.
On a personal note, I am an engineer, but I would never call myself a scientist. I consider myself a super-applied physicist. Engineers, and Computer Scientists for that matter, are concerned about applying theoretical advances to real world (or not so real world) problems.
As an example of this, many Comp Sci researchers have made real contributions to linear algebra, not so much for the sake of pure linear algebra, but for solving problems in graph theory, searching, graphics, et cetera. In fact, if it were up to me, I would split Comp Sci into two fields, Software Engineering, and Software Research.
For fuck's sake you guys. They are never ever ever ever going to have billions of these devices on one substrate. The lattice constants of the different components of the heterestructure strain the device and GREATLY lower yields. In my professonial opinion, yield is probably at 10s of devices now and will never reach more than a couple of thousand devices.
And yes, power will be high. 10+ Watts for a 100 xtr circuit is not surprising.
For God's sake man, they aren't going to be using these devices in a microprocessor pipeline. Ever. Read "Physics of Semiconductor Heterojunctions" by Sze and repent!
Here's a data point for you... 90nm CMOS transistors can run about about 40 GHz. There are no 20 GHz microprocessors. There are 2 GHz, microprocessors however. Therefore a 600 GHz microprocessor using these guys is never ever gonna happen.
And no one would ever design a circuit that would require a signal to traverse billions of these transistors. Come on now. We're smarter than that.
One thing which is not often mentioned, however, is that the yield on these kinds of III-V materials is low. What I mean is they probably would have a very difficult time getting 10 or so of these devices working on the same chip. Production Indium-Phosphide processes can get maybe a couple of 1000 devices working together. Quite a difference compared to billions of CMOS transistors. The main application of these guys, in case you care, is high performance radar receivers for the military, particularly electronic warfare receivers. Slightly less bleeding-edge devices of this type are used in high speed wireless and optical communications systems. And contrary to what other people might say on here, many of the applications are digital. In fact, 100+ GHz logic is a reality, just not for computing applications.
I'm very interested. What language do you think is good for today? Python? I mostly program in C, VHDL, and MATLAB but I'd like to learn a new general purpose language.
Standard CMOS logic power dissipation varies as the square of the supply voltage, and is linearly dependent on switching frequency. It has nothing to do with Si, it has to do with the structure of standar CMOS logic. Current mode logic, like ECL in Bipolar tech or CML in CMOS, has a more constant power dissipation, regardless of speed.
Lowering VDD (supply voltage) is the number one thing you can do to lower power diss, but there of course is a limit to that.
I work on CMOS analog chips and am interested in getting into III-V design (particularly data converters in Indium-Phosphide). Where do you work? Reply here are email me at carl_r_grace@yahoo.com
Oh we still have those deadly lawn darts. They are a commodity that I will always cherish. Of course, I did enjoy many other activities similar to what you describe. One of my favorite was dodgeball on bikes. My cracked teeth did not enjoy it so well though.
I played that game too when I was a kid. Did you ever play "Bike Jousting"? What started innocently enough as "Bike Polo" on a tennis court turned into an appallingly brutal game where tennis rackets, speed, and a taut net virtually ensured serious injury. Blood on the tennis court, literally.
And let's not forget the time my friend and I played "Kung Fu Theatre" on the front sidewalk. Thank god those teeth weren't premanent.
Interesting story, andreyw. I would say that 7th through 12th grade were the biggest waste of MY time ever. I too only learned about things on my own, through compulsive reading (anything I could get my hands on) and some computer programming. I'm very lucky I was able to survive in College, because I had no study skills to speak of (even with my above 4.0 grade average). It's funny, but I learned the most from my Cross Country and Track teams. Namely, hard work and self-respect. Not things you learn in an American public high school. At least in California.
Re:Obligatory Animal Rescue Link
on
Re-Pet a Reality
·
· Score: 1
If you look at the website, you will see that they pay spay/neuter clinics (like humane societies) for the eggs. That money is used to provide more spay and neuter services. So, the money IS spent on population control. It's also working toward commoditizing the process, which could mean great things for species near extinction.
Slashdot Mirror
Tapecutter,
No, the USA doesn't have that concept (mature age student) in any formal way. You are treated no differently from a new high school graduate. However, most "Mature Students" would go the Junior College route, which makes it significantly easier to get into a given state-supported school. I agree that older students are probably more likely to be motivated, my own personal observations supports that. And as for dropping out at 16, then getting back to school, and being successful all the while... may I only say "GOOD ON YA!"
Carl
He meant testicle feedback. It is a new gaming paradigm that involves a car battery and some electrodes or some-such...
He was also a very nice man. I knew him ten years ago when I worked at Lawrence Berkeley National Lab. The first time I met him was when he ran into me (literally) in the cafeteria and he said "excuse me son". This is amazing for a up-and-coming PhD student. Wow.
The DOE runs our system of national laboratories, and is the successor to the Atomic Energy Commission. They aren't all that concerned with gasoline, as that is a small part of their work. They mostly work on nuclear weapons, fusion research, high-energy physics, renewable resources, etc. I used to work at Lawrenece Berkeley National Lab designing subatomic particle detectors. I couldn't give a rats ass about how much you spend for gas.
A processor does do significant work. For example, the electrons in the channels of the transistors move against some finite resistance, phonons are released as electrons fall from the conduction band into the valence band, etc. It only doesn't produce light because there is a momentum change (hence the phonon) when the electrons do so. In direct bandgap semiconductors, light is released, hence LEDs and semiconductor lasers. The reason there is heat in any processor is the fact that unless there is some resistance somewhere, the clock rate could be infinite since it would imply that any capacitance could be charged up in zero time.
In ten years a few engineers will be designing new classes of electronics based on quantum principles.
As another replier pointed out, the silicon transistor is based on quantum principles. In particular, it's operation cannot really be understood without some knowledge of the Quantum Theory of Solids, which was originally proposed by Fermi, among others. Basically, only through statistical mechanics and other quantum stuff can you show that there are such things as "minority carriers" in semiconductors and that they exist in enough quantities to make a useful current. I think you meant that the theory of computing would be quantum, not the electronics with which it is realized. Also, lasers are based on photons. And we have memory devices in the lab based on spin (didn't you mean electronic spin rather than magnetic spin?)
Can you point us to a picture of this famous t-shirt? I'm curious.
This begs the question...
Begging the question means presupposing your conclusing, a form of circular argument. It does not mean a question that begs to be asked. Repent!
The Laser?
The laser was first demonstrated at Hughes Research Labs. www.hrl.com
So, if a guy wanted to build a more "modern" homebrew CPU, what options are there? Are there any decent CAD tools that don't cost a thousand million dollars? And once a layout is done, is there anywhere you can get just one single chip made for a reasonable price?
LesPaul,
I built a CPU using freeware tools as part of my PhD project. The paper is "A 12-bit 80MS/s Pipelined ADC with Bootstrapped Digital Calibration" published in the IEEE Journal of Solid-State Circuits, May 2005. You can google the title if you want to see the paper. Anyway, I designed a 24-bit microprogramed CPU in 0.25um CMOS to act as the calibration controller of the Analog-to-Digital Converter (ADC). The project was great because I got to design the architecutre of the CPU, the microcode/instruction set, I wrote a custom assembler, etc. I designed the circuits using viewdraw (public domain UNIX schematic capture tool). I designed the logic and tested the circuits using a public domain VHDL simulator (can't remember which one.. alliance maybe?). I laid out the circuit using Magic, a public domain layout editor running on Unix or Linux. The only thing that cost money was fabricating the chip. There is a service called MOSIS (www.mosis.org) that will do multi-project runs to lower the cost for you. I think the cheapest you will get is a couple of thousand bucks for 40 parts or so. Mine was something like $40k but I had high performance analog circuits in a fancy process. Email me if you need more info at carl.grace@yahoo.com
Cheers,
Carl
You think that is a lot to pay for gas? I paid 2.39 a gallon for unleaded today in California.
But see the problem is when the other companies have people "over a barrel" or are outsourcing, then they have an advantage over the one thats playing nicely with the union. And eventually the unionized company falls behind, more than likely because they can't provide as many incentives to investors so they can't grow like the non unionized companies.
That's what trade unions or industrial unions are for.
Of course (TANGENT ALERT!), the remaining reference to "natural phenomena" still precludes even the question "is computer science a science?"
EngrBohn: I've always thought that disciplines with the name "science" after them are never really a science in the strictest sense. They are more likely applied forms of another discipline, and while valuable, not a science in their own right. For example, Material Science is really applied solid-state physics, Political Science is applied sociology, and Computer Science is a mixture of applied mathematics and engineering.
On a personal note, I am an engineer, but I would never call myself a scientist. I consider myself a super-applied physicist. Engineers, and Computer Scientists for that matter, are concerned about applying theoretical advances to real world (or not so real world) problems.
As an example of this, many Comp Sci researchers have made real contributions to linear algebra, not so much for the sake of pure linear algebra, but for solving problems in graph theory, searching, graphics, et cetera. In fact, if it were up to me, I would split Comp Sci into two fields, Software Engineering, and Software Research.
Carl
Funniest thing I've seen on Slashdot in weeks. Bravo!
For fuck's sake you guys. They are never ever ever ever going to have billions of these devices on one substrate. The lattice constants of the different components of the heterestructure strain the device and GREATLY lower yields. In my professonial opinion, yield is probably at 10s of devices now and will never reach more than a couple of thousand devices.
And yes, power will be high. 10+ Watts for a 100 xtr circuit is not surprising.
For God's sake man, they aren't going to be using these devices in a microprocessor pipeline. Ever. Read "Physics of Semiconductor Heterojunctions" by Sze and repent!
Here's a data point for you... 90nm CMOS transistors can run about about 40 GHz. There are no 20 GHz microprocessors. There are 2 GHz, microprocessors however. Therefore a 600 GHz microprocessor using these guys is never ever gonna happen.
And no one would ever design a circuit that would require a signal to traverse billions of these transistors. Come on now. We're smarter than that.
One thing which is not often mentioned, however, is that the yield on these kinds of III-V materials is low. What I mean is they probably would have a very difficult time getting 10 or so of these devices working on the same chip. Production Indium-Phosphide processes can get maybe a couple of 1000 devices working together. Quite a difference compared to billions of CMOS transistors. The main application of these guys, in case you care, is high performance radar receivers for the military, particularly electronic warfare receivers. Slightly less bleeding-edge devices of this type are used in high speed wireless and optical communications systems. And contrary to what other people might say on here, many of the applications are digital. In fact, 100+ GHz logic is a reality, just not for computing applications.
I'm very interested. What language do you think is good for today? Python? I mostly program in C, VHDL, and MATLAB but I'd like to learn a new general purpose language.
Probably the funniest thing I've ever read on Slashdot. Here Here!
Isn't Pat Robertson the greatest combat marine ever?
To quantify this:
Standard CMOS logic power dissipation varies as the square of the supply voltage, and is linearly dependent on switching frequency. It has nothing to do with Si, it has to do with the structure of standar CMOS logic. Current mode logic, like ECL in Bipolar tech or CML in CMOS, has a more constant power dissipation, regardless of speed.
Lowering VDD (supply voltage) is the number one thing you can do to lower power diss, but there of course is a limit to that.
I work on CMOS analog chips and am interested in getting into III-V design (particularly data converters in Indium-Phosphide). Where do you work? Reply here are email me at carl_r_grace@yahoo.com
Oh we still have those deadly lawn darts. They are a commodity that I will always cherish. Of course, I did enjoy many other activities similar to what you describe. One of my favorite was dodgeball on bikes. My cracked teeth did not enjoy it so well though.
I played that game too when I was a kid. Did you ever play "Bike Jousting"? What started innocently enough as "Bike Polo" on a tennis court turned into an appallingly brutal game where tennis rackets, speed, and a taut net virtually ensured serious injury. Blood on the tennis court, literally.
And let's not forget the time my friend and I played "Kung Fu Theatre" on the front sidewalk. Thank god those teeth weren't premanent.
Interesting story, andreyw. I would say that 7th through 12th grade were the biggest waste of MY time ever. I too only learned about things on my own, through compulsive reading (anything I could get my hands on) and some computer programming. I'm very lucky I was able to survive in College, because I had no study skills to speak of (even with my above 4.0 grade average). It's funny, but I learned the most from my Cross Country and Track teams. Namely, hard work and self-respect. Not things you learn in an American public high school. At least in California.
If you look at the website, you will see that they pay spay/neuter clinics (like humane societies) for the eggs. That money is used to provide more spay and neuter services. So, the money IS spent on population control. It's also working toward commoditizing the process, which could mean great things for species near extinction.