The near future and electronics...as I see it.

HollyMinkowski

The thread on the propeller forum re Implementing
some sort of Parallax supported C somehow got me thinking
about 10-30 or so years ahead and what changes we will see
in electronics and our beloved micro controllers.

I'm always reading everything I can get hold of about the
future trends in this industry...this is what I foresee, what are your
thoughts?

---

In a decade or two things will change a lot. Controllers will be far more advanced
and enormously faster. I don't think that most of them will be little sm chips with
hundreds of pins, I think most will have optical I/O and need only power pins.
They will be easier for hobbyists to use than a dip chip is now. I can't see many
of them being single core devices, and the increased power will
finally enable great things to be done purely in high level languages.

Time passes quickly, we will soon move beyond 2d chip design and needing every
part to be hard wired together in a rats nest of connections....it will be good riddance
to multi layer boards. Just placing devices in close proximity will link them together.
Pretty much every part will be a smart part and can be modified at will. No more cables
with connector pins to connect projects with external stuff...just optical cables
with no metallic conductors at all...and these can be pretty long too. There is just
going to be a lot less copper used when creating electronic appliances. Of course
things like motors will need good conductors to connect to power but that is
all they will need, the rest will be optical, perhaps in some cases rf but I see
problems with too much rf being used.

An experimenters board with only power sockets will enable you to quickly
build up an incredibly complex device just by plugging in tiny parts and
creating the software to define and control it all. In many cases the only
copper wiring a project will have will be the 2 power leads. Write the code and stick a
touch/voice I/O visual display on a case and you are done.

I can't wait for it all to arrive! It will be so much cleaner and easier to deal with.
Projects created by selecting the proper smart parts blocks to assemble in close
proximity. Processors, multi-function rf blocks, network bocks..etc. Of course blocks
with multiple functions will be made, so you could choose a cheap one with everything
you need in some cases....a single part project.

Using physical conductors to carry anything other than power is just old fashioned IMO.
Everything is data (pretty much)

Ravenkallen

I don't know, I don't think optical tech will advance that much for small scale apps like micros. I do agree that the whole computer industry will move more toward modules that have a bunch of extras like, Wireless comms, built in sensors, audio output and input, increased program space. They might do that "smart interface " thing. It would probably be wireless, so all you would have to do is just bring the device in range and it could read and write from it without even hooking any thing up( Unless you shared the same power source). I wish they would make DIP micros with Megabytes of program space, instead of kilobytes. NOW that would be cool.

Unknown

As long as it is cheap and efficient, companies will pay for it.· A retired engineer told me that you need a fundamental chip because the top of the line might not be here tomorrow.··Chips with staying power will be here tomorrow.·

Tubular

Holly,

Fascinating to speculate. 10-30 years ahead is not far (its certainly easy to look back 15-20 years, and extrapolate 40 with "now" as a midpoint), so the seeds of the technology you're talking about are already well and truly sewn.

From my point of view I don't see the benefit in optical except for aggregated comms. As consumers we've developed a taste for wireless comms. If they ever get wireless power transfer really happening, we'll never go back to anything tethered, even if it has huge bandwidth like optical. Instead of optical I'd be tipping low power radio, with micros having that on-chip, and plastic circuits...

The whole printed electronics arena is really going places. Its going to be so much cheaper and more flexible than PCBs. It fits right in with micros using less and less power (you certainly don't need the conductivity of copper to signal data). If you want to get an idea sign up for the newsletter at www.idtechex.com/, or have a browse around the site

Whats the "problem with too much RF being used?" Security?

tubular

ElectricAye

Thinking more at the edge of 30 years here... I believe the next revolution in information processing will spin out of the world of biology. Biomolecules appear to be capable of doing some sort of supercomputing capability of some kind, possibly at the quantum level. I think science has just nicked the tip of this iceberg recently but the potential is possibly mindblowing. In fact my only concern is that its capabilities might somehow surpass our ability to really comprehend it, grasp it, and make it useful. We might have to invent some new kind of way of thinking just to deal with it. Our minds work mostly via metaphor, and if you don't have a good metaphor for a natural phenomenon, things are hard to grasp. Also, there's really nothing that says mathematics is the ultimate language of nature. We might have to tap into our artistic way of thinking or employ some unused chunk of brain matter or- whoknowswhat - just to derive some new angle of approach to understanding this process, a process that nature has had at least a billion years to tinker around with.


my 2 pennies worth

mctrivia

personally i do not think the optical or wireless fully modular system will happen. mainly because it would put to many people out of work and would require every device to be smart making parts more expensive.

However i do think that we will likely see a lot more PSOC systems where you will have a circuit board 1 chip on it with a trillion gates on it, and possibly an optical port to connect to a more powerful optical computer.

One thing I am pretty sure of 30 years from now computers will probably be about 1 billion times more powerful.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Lots of propeller based products in stock at affordable prices.

Unknown

ElectricAye said...
Thinking more at the edge of 30 years here... I believe the next revolution in information processing will spin out of the world of biology. Biomolecules appear to be capable of doing some sort of supercomputing capability of some kind, possibly at the quantum level.

Some colleges / universities are studying nanotechnology so I think that microcontrollers will go in a direction that a lot of start-up companies won't be able to compete with.· In economics, they teach that resources are limited and I·affirm it takes me minutes longer to commute to work each year because there are more and more people driving which means there will be less resources for every human being.· Therefore it would almost make sense (no one has a crystal ball) that microcontrollers may go the nanotechnology route to meet the demand since resources will be limited.

James Newman

The problem with rf is that there is only some much noise that we can throw up in the air and still receive. Even with the technology available... not sure it's useful anything 'critical'. (Like to see the wireless controls for a machine in a shop full of welders... that'd just be dandy...) For entertainment... this will continue to become more popular.

Fiber optics have been in use for a long long time. I see cnc controls that utilized them heavily from 15+ years ago. They will continue to increase in use due to the bandwidth, speed, latency, and lack of emi. IMO data traveling over any distance longer than a foot or two should be fiber optic and serial.

Things can't get outside of our grasp of thinking... because we design them. Everything that has always been possible and is only made a reality by our realization of their workings.

Unknown

mctrivia said...
One thing I am pretty sure of 30 years from now computers will probably be about 1 billion times more powerful.

Microsoft had the browser wars and then it was the online search wars.· I felt like they had this Pinky and the Brain complex where the Brain tries to take over the world.· They tried to sue Linux developers.· They entered the game console industry.· It will only be a matter of time when they enter the Microcontroller world because they are trying to take over the world, because they want Windows running on every device.· That is why they have wanted to make the personal computer look more like a TV.· That is why they applied for a software patent on double clicking on the mouse but an EU judge said,'no way'.· Don't think it is too far fetched because for a while they were talking about home automation which is what microcontrollers do.

erco

Sure, high end stuff will continue its amazing progress. But the average hobbyist can only learn, do, and solder so much. The Stamp 1 came out in 1994; it and its derivatives are going strong 16 years later. Tracy Allen did "everything possible" with the Stamp over 10 years ago, and there are plenty of us just scratching the surface of what he did long ago. Someone recently posted about how they got someone hooked on a Stamp just flashing LEDs. Some people only want to go so far and spend so much. I think there will always be a need for some simple introductory chips, something like the Stamp, Arduino, etc, even 20 years from now. Unless the Atomic Flame Deluge comes before then and the Great Simplification begins. (Obscure "Canticle for Liebowitz" reference, great book...)

BTW, Holly: Physical conductors for mere power is already passe. The new rage is proximity charging, just throw your camera & cell phone on an RF mat, they recharge & download/upload wirelessly. It's all hifi wifi bluetooth Xbee Zigbee hidef mp5 4D blackfin web 2.0 firewire cable-able compatible. Surely you read all about it in the newspaper...

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·"If you build it, they will come."

bill190

HP's possible future "memristor" product could change things!
http://www.dailytech.com/HP+Labs+Shows+New+Memristor+Tech/article18092.htm

http://en.wikipedia.org/wiki/Memristor

kwinn

I agree with some of your conclusions for the future, but not all of them.

>> In a decade or two things will change a lot. Controllers will be far more advanced and enormously faster.

No doubt there will be faster chips with more memory, possibly even 64 bit cpu's and optical communications, but there will still be a need for a lot of small 8 bit cpu's. After all, how much processing power do you need for a thermostat, toaster timer, or similar application.

>> I don't think that most of them will be little sm chips with hundreds of pins, I think most will have optical I/O and need only power pins.

While it is possible that an optical communications bus will replace the copper based serial buses in use today, the main function of microcontrollers is controlling power to I/O devices so I am pretty sure the pins will be with us for quite a while.

>> They will be easier for hobbyists to use than a dip chip is now. I can't see many of them being single core devices, and the increased
>> power will finally enable great things to be done purely in high level languages.

No arguments here. The propeller chip with it's 8 cogs, built in boot loader and spin interpreter is the way to go.

>> Time passes quickly, we will soon move beyond 2d chip design and needing every part to be hard wired together in a rats nest of
>> connections....it will be good riddance to multi layer boards. Just placing devices in close proximity will link them together.

As we get more transistors on the chips the number of chips required to accomplish a function will decrease but the chips will still need to be connected and powered somehow. Chip to chip signaling may be optical but they will still need conductors for power so I think PC boards will be around for a while yet.

>> Pretty much every part will be a smart part and can be modified at will. No more cables with connector pins to connect projects with
>> external stuff...just optical cables with no metallic conductors at all...and these can be pretty long too. There is just going to be a lot less
>> copper used when creating electronic appliances. Of course things like motors will need good conductors to connect to power but that is all
>> they will need, the rest will be optical, perhaps in some cases rf but I see problems with too much rf being used.

There will be "smart parts" consisting of an integrated processor and power control circuit connected to a bus (optical or copper). I think this will be more in the nature of an evolution of programmable logic controllers. Instead of being mounted on a DIN rail the smart part will be distributed close to the device it controls and either powered from the same source as the device or a pair of power conductors.

>> An experimenters board with only power sockets will enable you to quickly build up an incredibly complex device just by plugging in tiny
>> parts and creating the software to define and control it all. In many cases the only copper wiring a project will have will be the 2 power
>> leads. Write the code and stick a touch/voice I/O visual display on a case and you are done.

>> I can't wait for it all to arrive! It will be so much cleaner and easier to deal with. Projects created by selecting the proper smart parts
>> blocks to assemble in close proximity. Processors, multi-function rf blocks, network bocks..etc. Of course blocks with multiple functions will
>> be made, so you could choose a cheap one with everything you need in some cases....a single part project.
>> Using physical conductors to carry anything other than power is just old fashioned IMO. Everything is data (pretty much)

This would all be great if it could be done, but it has some major hurdles to overcome. How do the blocks communicate? RF is subject to interference, and the problem gets worse as the number of blocks increases. Optical requires that blocks be aligned for line of sight signaling, requires light guides for non line of sight, and may also be subject to interference from external light sources.

The biggest hurdle of all may be getting all the chip makers to agree on a communications protocol/standard.

Unknown

Four Key Trends In Microcontrollers

Reduce Power Consumption
Increased processing power
Smaller form factors
Faster time to markets
More analog and overall integration

Cluso99

I read an Intel article in 1991 predicting the computer chip of 2001 (IIRC the dates). I did find the article later and it was way off in many respects - they said 4 processors and on chip memory. One of the processors was graphics. Well we are not really there yet although we now have multiple, but identical, processors on Intel chips.

One thing is for sure, the number of transistors on a chip will still increase at least at the same rate, probably using different technoligies though. The chips will become multilayer (they are now but not really 3D yet).

I have seen chips go from ~50 transistors (in my teens) to > 2 billion in the latest Intel chip. In 10 years it will be about 200 Billion. However, I bet you cannot predict what functions will be in the main chip in production in 10 years.

Now what do you think Chip could do with 2 billion transistors? Nothing because he cannot afford to even design it :-(

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:

· Home of the MultiBladeProps: TriBlade,·RamBlade,·SixBlade, website
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: CPUs Z80 etc; Micros Altair etc;· Terminals·VT100 etc; (Index) ZiCog (Z80) , MoCog (6809)·
· Prop OS: SphinxOS·, PropDos , PropCmd··· Search the Propeller forums·(uses advanced Google search)
My cruising website is: ·www.bluemagic.biz·· MultiBlade Props: www.cluso.bluemagic.biz

mikediv

Well I will be dead, but you guys have some very insightful views. I think one of the things that we will see for sure is a race for multicore processors that and a race for as many processors as can be fitted on one chip.

Then and this is an absolute certainty to cripple the enormous advances that will take place Microsoft will develop windows into a huge and disgusting bloated OS that will need 50 Terabytes of Ram and a few billion terabytes of storage
to give us a system that will run slower than my 32K CPM 8085 system . Take 40 minutes to boot crash on a regular basis require service packs and daily updates (adding to its size) and for some reason run slower and slower the longer you run it
So basically evening out any performance gains multi processor or smart systems advantage to that of a 120 year old 8 bit 6502,LOL Well I say some of this in jest but like the saying goes "there is many truths spoken in jest"

K2

Until it is less expensive to employ a wireless link between two chips on the same board than it is to use simple copper conductors to connect them, we will continue to use PCB traces to connect chips. Of course the trend all-along has been to put more and more functionality on a single chip. This results in fewer chips being needed And that means fewer interconnectionsns are required. Compare, for example, the Apple II PCB to the Propeller stick. That's the ongoing trend I'm talking about.

Meanwhile, consider how sophisticated a wireless link would need to be to transfer data at the same rate as a simple whisker of copper is able to do so. There is simply no comparison in terms of cost, complexity, size, reliability, immunity to interference, power consumption, etc, etc.

skylight

In the future it may be a case of being familiar with chemistry as opposed to electronics, we may be constructing living circuits and adjusting their capabilities by adding chemicals, even genetically.
The need for ever more sophisticated processing may drive the way from PCB's to organic depositories that grow and adapt to the tasks given it?
Instead of working out currents it would be a case of how many drops.