No, microwaves are not going to destroy a tiny, Si-on-plastic chip. The wavelengths are too long, even at 4GHz, to couple much power into the chip
Yes, but since this one is a RF-Tag it also got an external antenna. And for some (not that surprising) coincidence the hitachi chip works on the 2.5GHz band which is also the frequency used in microwaves.
Well - but in the 80ies and before the complexity and power consumption of clock distribution networks was neglectable. Today a two digit percentage of space and up to 40% of the power is consumed purely by routing the clock signals.
So the extra complecity added by async logic becomes more and more feasable. However I dont think we will see fully async chips in widespread use, soon. Moreover bigger and bigger parts of the chip will work async. with a synchronous frontend.
Capacitor leakage might be a problem, but it's not impossible to translate those old designs into static CMOS. And the patents have expired!
Well, i dont think the patents are the problem. I guess today it is possible to build a more efficient CPU with the same amount of transistors without having to rely on i4004 patents.
The smallest elements you can print are about 25 micrometers (.001 inch), which is over 100 times the linear dimensions in a modern Pentium chip. And that
The i4004 was manufactured in a 10micrometer process. So one could maybe do a i4004 clone with just 2.5^2=6.25 times the size of a real i4004.:)
Ok, probably it would not work due to capacitor leakage (it used dynamic registers) and if it worked it could just be clocked at xx kHz..
While I think this is a very interesting technique I totally dislike the hype coming with it. Maybe it is actually possible to manufacture cheap electronics with a cheap printerlike machine in a few decades, but this will be limited to slow low complexity circuits, like electronic price signs etc.
And even IF it will be possible to manufacture higher end electronics this way - it will most probably always be cheaper to do it industrial. Just compare it to books. Today everyone is capable of printing his own books with his very own printer - but how many people are actually doing it ? It is much more convenient to buy books.
Sega Dreamcast.
Nope.. the dreamcast uses a SuperH Core.
Re:Yeah but, chip making isn't as easy as writing
on
Open-Source Processors
·
· Score: 2
Yeah, if you want gigantic, slow, hot, wasteful chips. If you want high-performace chips, like we're talking about here, you would at least have to design the datapath by hand. And have you ever synthesized memory, by chance? Probably not. I bet you would say "just get a memory compiler". What about clock generation and distribution? What about high speed digital signaling?
Motorolas new coldfire is 100% Synthesized, including the on chip cache. Of course they have probably helped the router a bit while placing the cache blocks and clock distribution. But the core byself is a sea of gates.
One thing that is nice is that they use a LOT less power than conventional LCD's. Also, there is no ghost when animation occurs like in conventional LCD displays..
Do they really use less power than reflective LCDs ? I somehow doubt this. Probably they use a lot less power than LCD + Backlight, which is a big difference.
Collossus was not used to decypher enigma codes.
This had already been done by some polish mathematicians, and later by Turing and his "bomb". (Electromechanical decyphering device)
The collossus was used to decypher encrypted teletext transmissions. The transmissions were encrypted with pseudorandom numbers. Due to a mistake during one transmission by the germans (they send the same message twice) the english were able to figure out the generator polynom. The collosus was used to correlate encrypted transmission with this generator polynom.
Of course a 100GHz++ CPU cant work with a clock distribution network like current cpus. Other techniques have to be used.
Some buzzwords: Wave Pipelining, Fully Asynchronous designs, Systolic Arrays.
Please check the link for the 770GHz flip-flop. They are already working on a CPU thats supposed to reach 100GHz. And they already have to use wave pipelining etc.
If current trends are projected forward, by 2020 a bit of memory will be a single electron transistor, traces will be one molecule wide, and the cost of the fabrication plant will be the GNP of the planet. The speed of light imposes practical limits on how large you can make a chip and how fast you can clock one. This is why we'll have GHz chips, but fundamental physical laws prevent THz chips.
The current speed record for a digital flip-flop is 770GHz..
While this technique is nowhere close to go into mainstream (or even scientific) computing it still shows that circuits operating in close-to-one-THz range are possible. Things might be different in twenty years. And Prozessors in the THz range would for sure be nowhere close to the CPUs we have today. Probably heaviliy asynchronous processors using architectures like systolic arrays etc. have to be used.
Also dont forget that mainstream CPUs are not made with the fastest technology available, but with the cheapest. By the use of GaAs Cray was able to achieve clocking speeds of around 1GHz in a time when a stock PC was clocked at 33MHz - so what might a GaAs CPU with current technology scale up to today ? Or how about BiCMOS ? (given that you have a personal power plant and some insane cooling device;) )
The article claims the that GFX chip was developed
by ATI. I always thought ArtX (www.artx.com) was deceloping the gfx chip ? Probably a bad typo.
Btw. it looks like Nintendo beats Sony again when
it comes to manufacturing costs. Only two bigger chips required for the full console! (Though the production costs will probably be quite high in the beginning due to the huge amount of embedded dram)
Well - there are also many situations were you just cant use an OS on your MCU and were 32 bit is just a waste. Think if time critical state machines, applications that were done with discrete TTL chip in the past etc. It will still take years until 16 or even 32 bit controllers are as cheap and versatile as 4 and 8 bit ones.
Actually I dont think there are that many situations where it would really make sense to replace an 8 bit mcu with a linux bloat system. (sorry:)) An appropriate way would be to use an additional embedded CPU with an OS for communications tasks, whenever this is required.
The Commodore PET is oft-forgotten, and I hope that some of us remember it. It had a steel case, an internal cassette drive, and a 9" display. It even shipped with 8K of RAM, which was a lot in those days. Chuck Peddle was a god. He later invented the Motorola 68000, as I recall (I may be wrong. Corrections gratefully accepted).
Actually it was the 6800. He developed the 6800 even before the 6502. He (and his partners) planned to do a low cost CPU which was bus-compatible to the 6800 called 6500. Appearantly Motorola sued the sucessfully and they had to change the bus a little and called the CPU 6502.
Slashdot Mirror
No, microwaves are not going to destroy a tiny, Si-on-plastic chip. The wavelengths are too long, even at 4GHz, to couple much power into the chip
Yes, but since this one is a RF-Tag it also got an external antenna. And for some (not that surprising) coincidence the hitachi chip works on the 2.5GHz band which is also the frequency used in microwaves.
The Chip will get roasted badly.
Wilhelm Siemens was the brother of the Siemens founder, Wernher von Siemens, and was responsible for the English subsidiary.
Further information here: click
Most probably you are referring to blue SiC LEDs. These are however not related to the article.
So the extra complecity added by async logic becomes more and more feasable. However I dont think we will see fully async chips in widespread use, soon. Moreover bigger and bigger parts of the chip will work async. with a synchronous frontend.
750kHz exactly :)
Capacitor leakage might be a problem, but it's not impossible to translate those old designs into static CMOS. And the patents have expired!
Well, i dont think the patents are the problem. I guess today it is possible to build a more efficient CPU with the same amount of transistors without having to rely on i4004 patents.
The i4004 was manufactured in a 10micrometer process. So one could maybe do a i4004 clone with just 2.5^2=6.25 times the size of a real i4004. :)
Ok, probably it would not work due to capacitor leakage (it used dynamic registers) and if it worked it could just be clocked at xx kHz ..
While I think this is a very interesting technique I totally dislike the hype coming with it. Maybe it is actually possible to manufacture cheap electronics with a cheap printerlike machine in a few decades, but this will be limited to slow low complexity circuits, like electronic price signs etc.
And even IF it will be possible to manufacture higher end electronics this way - it will most probably always be cheaper to do it industrial. Just compare it to books. Today everyone is capable of printing his own books with his very own printer - but how many people are actually doing it ? It is much more convenient to buy books.
Sega Dreamcast.
Nope.. the dreamcast uses a SuperH Core.
Motorolas new coldfire is 100% Synthesized, including the on chip cache. Of course they have probably helped the router a bit while placing the cache blocks and clock distribution. But the core byself is a sea of gates.
Well.. actually the design of the original beetle was done by Ferdinand Porsche, who later founded Porsche.
And in addition it took another 30 years until someone was able to describe how the crystal actually worked! (Walter Schottky, 1938)
I doubt it.. In Hamburg/Germany one radio station has a 810+x title user voted music countdown each year. Takes three days.
Wow.. 35 full shares. Pocketmoney trader, eh ? :)
Do they really use less power than reflective LCDs ? I somehow doubt this. Probably they use a lot less power than LCD + Backlight, which is a big difference.
Well, I can stream the videos :)
http://www.home-micros.fre ese rve.co.uk/zx80/zx80.html
Oh well, I am not sure whether it was really a LFSR PRN generator, but at least something similar.
Collossus was not used to decypher enigma codes.
This had already been done by some polish mathematicians, and later by Turing and his "bomb". (Electromechanical decyphering device)
The collossus was used to decypher encrypted teletext transmissions. The transmissions were encrypted with pseudorandom numbers. Due to a mistake during one transmission by the germans (they send the same message twice) the english were able to figure out the generator polynom. The collosus was used to correlate encrypted transmission with this generator polynom.
Some buzzwords: Wave Pipelining, Fully Asynchronous designs, Systolic Arrays.
Please check the link for the 770GHz flip-flop. They are already working on a CPU thats supposed to reach 100GHz. And they already have to use wave pipelining etc.
The current speed record for a digital flip-flop is 770GHz..
While this technique is nowhere close to go into mainstream (or even scientific) computing it still shows that circuits operating in close-to-one-THz range are possible. Things might be different in twenty years. And Prozessors in the THz range would for sure be nowhere close to the CPUs we have today. Probably heaviliy asynchronous processors using architectures like systolic arrays etc. have to be used.
Also dont forget that mainstream CPUs are not made with the fastest technology available, but with the cheapest. By the use of GaAs Cray was able to achieve clocking speeds of around 1GHz in a time when a stock PC was clocked at 33MHz - so what might a GaAs CPU with current technology scale up to today ? Or how about BiCMOS ? (given that you have a personal power plant and some insane cooling device ;) )
The article claims the that GFX chip was developed
by ATI. I always thought ArtX (www.artx.com) was deceloping the gfx chip ? Probably a bad typo.
Btw. it looks like Nintendo beats Sony again when
it comes to manufacturing costs. Only two bigger chips required for the full console! (Though the production costs will probably be quite high in the beginning due to the huge amount of embedded dram)
Are you sure the 4Mb version is already in production ? AFAIK Fujitsu is just about to sample a 1Mbit parts, but thats it.
I got a sample of Ramtrons FRAM. Its really neat - much more convenient than FRAM or EEPROMS. (say speed, paging, durability etc etc..)
Other nonvolatile RAM technologies:
Flash backed up sram with on die CAP:
(available now)
www.sram.com, www.zmdusa.com and more..
C-Ram, OUM or whatever it is called:
(appearantly not in production or sample
state, yet)
www.http://www.ovonyx.com/
Well - we already had this topic quite a while
ago.
50 GHz ? Check this out:
http://pbunyk.physics.sunysb.edu/~paul/uprocess
They also made a 770GHz flip flop.
Oh well - maybe the liquid helium cooling might prevent this from entering the mainstream market.
Well - there are also many situations were you just cant use an OS on your MCU and were 32 bit is just a waste. Think if time critical state machines, applications that were done with discrete TTL chip in the past etc. It will still take years until 16 or even 32 bit controllers are as cheap and versatile as 4 and 8 bit ones.
:)) An appropriate way would be to use an additional embedded CPU with an OS for communications tasks, whenever this is required.
Actually I dont think there are that many situations where it would really make sense to replace an 8 bit mcu with a linux bloat system. (sorry
Actually it was the 6800. He developed the 6800 even before the 6502. He (and his partners) planned to do a low cost CPU which was bus-compatible to the 6800 called 6500. Appearantly Motorola sued the sucessfully and they had to change the bus a little and called the CPU 6502.