Cluso99 said...
BradC:
I used to have the circuits of a mini-computer built in 74 series TTL. The processor was a dozen pcbs about 15" square and a wire wrapped motherboard. Then the memory - that was in blocks of about 6 pcbs of core memory 10Kx6bits. The terminal I/O channels were 2 pcbs per channel (max 20). Then the hard disk and mag tape took another 20 board cage. Don't ask too loudly or I might find you the circuits !!! The processor is remarkably similar to the prop.
I believe I could emulate this mini on one Blade #2 of my TriBladeProp, with the exception of the 200 maximum (typically 5-30) terminals and printers I will need a few props for this. FYI I have done the emulation on a 486, just for (stupid???) fun because I could, after which I had two potential customers, and one of which actually validated the code. A company in the UK did this and they sold heaps !!!
Now you have modified your code to compile our nutty projects, have you joined our fruitcake club ?*?*? lol
I have a "thing" for old tech... but SWMBO likes things clean, tidy and neat so I don't get to hoard like I used to... guitars and motorbikes I can get away with, but old mini's and tty's are right out.. If I ever find my old Apple ][noparse][[/noparse] though she'll just have to cope!
So, yes, I might be a bit nutty. I've been meaning to add the #define and other stuff to the code for a while but I got side tracked by an IDE, Syntax highlighting and other fun stuff I'd never really thought about playing with. I've been working on #include for a while also, but I'm not going to do it until I can come up with a neat way of incorporating the line numbers and filenames properly so I can cross reference the list file and compiler errors...
Emulation is fun.. I can see the attraction of it.. I just need a few extra hours in my day [noparse]:)[/noparse] You can thank a nasty stomach wog and the resulting night of insomnia for the #define processing though :O. Imagine how much I could get done if I got a bout of H1N1!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
And core memory is edging it's way back in as MRAM. Just have to fix those reset circuits, or there will be new meaning to the phrase "My computer crashed".
Video generator to set the address would take a few instructions (3?) but would wait until the bits are consumed by the waitvid. Also, depending on the number of bits and the clock rate, a relatively large amount of time would pass before the bits are set. See MagIO2's example of how this could be done http://forums.parallax.com/forums/default.aspx?f=25&m=348109&g=348221#m348221
Back in the day (where "the day" should be read as "the early 80's") I designed an entire CPU around 74LSxx logic, all the way through writing microcode for it, but I never built it. At that time it was actually an attractive idea because most CPU's were MOS or CMOS and limited to 1-2 MHz, while 74LSxx could run to 40 MHz, which in that day was considered a really fast clock. (The infamous C-shaped Cray supercomputer ran at 40 MHz.) However, right around the time I was drawing up my design (on an actual drafting board with pencils) clock rates and bus widths started to creep up in LSI, and by the time I had the money to do it it was clear I'd be creating an unsupported albatross that would soon be obsolete. It was an educational experience but I'm glad I didn't go through the effort of actualizing it.
A crazy thought did occur to me some months ago: Build an ALU from 74xx. Build some registers, Accumulator, Program counter, Stack pointer or just general purpose in 74xx. Make them 16 or 32 bit. Build some interconnecting MUXes to tie these together with a memory bus interface in the architecture of your choice. So far relatively easy to do.
Now the micro-code and control unit. That's a pain, big and hard to debug. So, you guessed it, use a Prop ! It's got enough I/O pins to drive the ALU commands, MUX directions etc. You can download the micro-code easily, single step through the micro-ops etc.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
Did you finish off your Prop CPLD board?
I'd like to explore the possibilities/capabilities of a Prop tied to either an FPGA or CPLD.
Seems to me an FPGA/CPLD that could handle all the I/O subsystems and memory management coupled with a Prop would make a nice and accessible dev system.
Coley
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ PropGFX - The home of the Hybrid Development System and PropGFX Lite
Still OT [noparse]:)[/noparse]
I have a couple or Prop/CPLD/SRAM/SDRAM designs in layout on a 3x4 FAB. The CPLD is Altera 7064 PLCC44 for SRAM and 7xxx PLCC84 for SDRAM. I chose PLCC since BatchPCB will produce >= 8/8 trace/space boards. The devices I fear are are too small for generic CPLD experiments though. How big a device do you think is practical for experiments?
We are all craving more memory. We have done this using only 12 I/O pins, leaving 20 I/O pins for development. Dr. Gouge has now created a way to link unlimited numbers of Propeller Proto boards together, each with the ability to daisy-chain up to 16MB of static ram, giving you an unlimited supply of COGs to run programs, servos, etc. with as much memory as you need to do any kind of project imaginable.
The wirewrap is only a prototype. It will be stamped on a PCB.
I hope you will be able to see the value inherent in a design like this. Dr. Gouge has made it possible for you to do things you never thought you would be able to do with a set of Propeller chips.
Just as a personal opinion ... Go easy on the sales pitch here. So far you've provided a lot of "this is the greatest thing since sliced bread" stuff and not much in terms of specific details. That's fine. It's all well and good to be excited about something you've spent a lot of time and effort and imagination on, but this forum has a variety of members, some with little experience and some who have spent years and years working in this area and who understand that there are many ways to do memory expansion for something like a Propeller and there are tradeoffs, like I/O pins used vs. speed of access vs. type of access (purely sequential vs. random) and there's no magic way around this.
Success with your new product will depend much more on support than anything else. It's easy to build a physically nice looking product. It's always a lot of work to provide documentation that helps people and sample software and I/O drivers that fit with what most people need to do or can be easily modified by and for those whose needs are different than most.
jazzed: "How big a device do you think is practical for experiments?"
How long is a piece of string? No really I have no idea. The concept was to have a CPLD help with the access from Prop to RAM whilst at the same time providing a bunch of I/Os to replace the Prop ones lost in driving all this. In the meanwhile Cluso came up with his TriBlade external RAM system using a 10 cent helper micro controller which is going to keep me busy for a while.
Just now I'm to snowed under by "things to do" I really must resist any more shiny things. I've promised myself to one day get into FPGA/CPLD and VHDL but when I do I want the biggest FPGA I can reasonably afford. I want to swim in a sea of gates for a while[noparse]:)[/noparse] and I know there are lots of "shiny things" in the FPGA world I will want to try out.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
Mike basically said what I was going to; there's no magic solution, no one perfect solution fits all, it's low I/O count and slow speed, high I/O count and fast speed, or a compromise somewhere in between. Plus the third side of the triangle, cost.
While memory is useful, particularly for data storage, when looking to extend program code space there's a lot more to it than just bolting some memory onto the Propeller.
I'm all for pushing boundaries, as a personal or technical challenge, to see if one can, or just for fun, but if I needed megabytes of high-speed memory I would probably be looking at something which could support that natively rather than jumping through the hoops which a Propeller demands. For $800 I could buy an Atom PC with 2GB of Ram, 160GB of hard disk and all else that a PC gives me, and still have plenty spare to buy all the Propellers I need for I/O interfacing.
Or maybe I'd just rip the innards out of a $10 router. Hmm; imagine how capable 80 would be when networked together over their inbuilt 100Mbps LAN
So here's the plan:
1) ARM or similar 32 bit CPU the cheapest we can find.
2) Some 10s of megabytes of RAM and FLASH for it.
3) It runs Linux.
4) A Prop or two with all Prop Demo board features.
5) All on the same board. No bigger than a little Eurocard.
6) Perhaps some Ethernet for the ARM
Items 1 to 3 gives the compute engine, memory, and networking.
Items 4) is there for video out and interfacing all those hardware gizmos we like to play with.
Anyone any idea how simply/cheaply this could be put together? I'm somewhat out of touch with ARMs and such but I have worked with boards like this with ATMEL chips that did not seem to be mindbendingly huge complicated projects to tackle.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
heater, that ARM/Propeller combo is the direction I wanted to go and started laying out an LPC version ... (staying under 3x4 is challenging). But then I keep thinking the PC thingy like Hippy mentioned would be easier to do and wouldn't cost too much. But then I wouldn't be adding to my skill set [noparse]:)[/noparse] So many things to do so little time. BTW: I've already written and tested the verilog for the 8 bit synchronous SRAM CPLD 7064 controller 3x4 FAB ... it's just a matter of deciding to spin the board.
It seems to me an experimenter's FPGA board should have at least a 64K macro cell Cyclone II or equivalent.
mallred, as far as I'm concerned you can sell anything you want if you can find a market for it in anyway you like. All you have to do is find paying customers who really believe in your value proposition ... or just like collecting things [noparse]:)[/noparse]
I think some of you are missing the point about WHY Dr. Gouge created this memory expansion board. There are a lot of valid criticisms about need. This is great, because what was created came out of a need that was not available anywhere.
Before Dr. Gouge started this project, his first question was, looking at the hobby robotics market, "Where is the intelligence?"
That one question started this whole project.
Dr. Gouge has developed software that can mimic the human brain. How many software programs out there do you know of that can do this?
His genious is in taking the neurobiological model of the brain, and transcribing that into code that approximates the brain's ability to learn new information and attach to old information such that real learning occurs.
His software has been proven in several military applications. His desire is to now bring this technology to the masses, or to researchers who will continue his work when he is gone.
When he started, he looked around for a platform that would support his technology. Nothing out there was able to support it.
He asked Parallax for a debugger. There was none. So he wrote one.
He built this system out of necessity, because no platform was out there that would do this. Dr. Gouge will be building the first cognizant android the world has ever seen. So, yes, we have an agenda.
What you seem to be missing is that the Parts and Pieces that get developed along the way have far reaching applications beyond his original need for a new platform.
Who knows but what others will do with the peices Dr. Gouge creates? This is what is exciting. The possibilities are endless. We are taking the Propeller and we are doing something large and exciting with it. The pieces that get created stand on their own merits as tools and products others can use and extend.
Don't miss the forest for the trees. There is something bigger happening here than you realize.
Considering that we still don't know how the brain encodes information, your claims for Dr. Gouge's ideas are a bit far fetched. There have been many learning systems developed over the years based on the then current understanding of the brain and cognition. Many of them (like neural nets) are in widespread use today although they're not anywhere as useful as was claimed at the time. We do know a lot more about the brain than 10 years ago or 20 years ago, but still have a long way to go.
Go easy on the "wild claims". They've been made over and over before and they're not helpful.
For a glimpse of Dr. Gouge's technology, (no, not neural nets, although he studied them for many years and decided that was not the approach to take) see the July/August issue of Robot magazine. He has an in-depth interview that I think you will find very interesting.
The "wild claims" he has proposed are helping to protect you and me as we speak. This is proven technology in military circles.
Ok, I will give you a hint.··You are correct.· We do not know how the brain encodes information.· Dr. Gouge's technology even mimics this.· He has figured out a way to mimic the brain WITHOUT having to know how the brain encodes information.· See the Robot magazine article for more...
A COG running at 80MHz executes code at 20MIPS. If it has to go to external RAM to fetch code (LMM) we're down to, what, 5MIPS. If it has to emulate some 16 bit CPU architecture (Hinted at here or on the web site) we're down to say 1MIP. If all 8 COGs are trying to run said code in parallel we are crawling.
So whatever brain emulating software there may be could for sure be run on my run of the mill dual core 3G PC with 3GIG RAM a lot cheaper than buying a truck load of Props to do it.
Also, is it really possible that some one has perfected world shattering human brain emulation software to run on a platform that does not exist yet? Surely it has taken years to perfect and is available, for show at least, on a PC.
I'm prepared to be surprised but it's going to be hard work.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
I can tell you that it has been done on the PC first and is now in operation in military robotics projects. We want to create an android, and therefore needed a mobile platform to do it. Dr. Gouge chose to implement this on the Propeller.
And yes, over 40 years of Dr. Gouge's life has been pursuing this one goal, machine intelligence.
We started the project in January. Give us some time. In four months we have developed the KISS Debugger, the SRAM Memory Expansion project, the skeleton for an operating system for the propeller we are calling the KISSOS, and are currently working on voice recognition software to eliminate the keyboard. Whew! We have not even come up for a breather. Give Dr. Gouge a little credit. He has been very busy :P
OK you need a mobile platform. Still there are many better cheaper ways to implement a massively parallel system than a crate full of Props (I'm prepared to be proved wrong here). Access to RAM is one issue. Communication between processing nodes is another (neurons tend to do a lot of that). Sharing RAM to communicate is one way but slows everyone down. Point to point links is another. The Transputer had that covered twenty years ago ago and the XMOS chips are picking up the flag now.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
I'm quite certain that the technology behind your company's RAM is sound, although extraordinary claims do require extraordinary proof. But, like two jigsaw pieces that don't quite fit together, your approach to the Propeller community is producing a mismatch that's resulting in some of the push-back you see here. If you'll permit me some constructive criticism:
1. Selling schematics and Propeller software just doesn't cut it with this crowd. The added value that produces sales has to come from well-crafted hardware. Kits that include a PCB (no wire-wrap stuff, please) and through-hole parts are okay; compact, completely assembled boards using SMT components are even better. What people are willing to pay for is the hardware that implements the IP. Hoarding the IP or trying to ration it by offering it for sale is not going to be a recipe for success. Basically, what this means is taking a device (which you apparently already have) and creating a product out of it. This requires a lot of extra work; but, to be taken seriously, you have to do it.
2. Your company's website seems like an attempt to create a cult of personality around Dr. Gouge, complete with photos of ID badges from his college days and prior employers. I don't quite get that, frankly. Nor will a majority of your potential customers, I'm afraid. This perception is only compounded by having you as a mouthpiece, while the mysterious Dr. Gouge remains cloistered somewhere "behind the curtain". I'm sure he has a very busy schedule, but it would really be nice to hear from "the man" himself.
I don't mean this to sound harsh or mean-spirited. On the contrary, I hope your business succeeds wildly. So I thought it would be useful for you to see how your company is being perceived. The successful businesses in this century are based upon openness, the sharing of information, and producing sound products backed by great customer service. If you want to see a perfect example of this, take a look around Parallax's website and at the things they sell, but especially at the incredible abundance stuff that can be downloaded for free — including programs from the Propeller Object Exchange.
The reason for the Propeller Proto board is so that the hobbyist could start small and grow to the desired or needed size. We are trying to be accomodating to the inventor/researcher/hobbyist on a relatively small budget.
His software will run on a 4MB expansion board. Memory represents neuronal links, and therefore, if you need your project to learn more, all you do is add memory.
We hope this approach will enable many more people to become involved.
Thank you for the input. I will definitely take it under consideration. We want this to be a successful venture, of course. If I have come across the wrong way, I apologize.
I feel we have something of value to offer the world. In four months, it is not yet polished, but we have the beginnings of something very special. My hope is to convey that in this forum, many of whom are potential customers.
The photos of Dr. Gouge on the site is because a magazine questioned his credentials, so I was asked to put those up for the world to see.
I hope to convey our desire to share this technology with the world and to create a successful business model around it. I know that service will be paramount and I pledge here and now to offer the best service possible.
So far Google has failed to turn up anything, for me, on him or his extensive cognitive research apart from what is in the machineinteltech web site.
Those pictures and such don't help, all we learn there is the he has once been a field service and customer engineer. They just deepen our curiosity....
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
On our About Us page we describe and link to some of his many patents. You might start there. If you notice, they link directly to the US Patent Office website. Also, we have posted several of his previous articles from various magazines.
Comments
I have a "thing" for old tech... but SWMBO likes things clean, tidy and neat so I don't get to hoard like I used to... guitars and motorbikes I can get away with, but old mini's and tty's are right out.. If I ever find my old Apple ][noparse][[/noparse] though she'll just have to cope!
So, yes, I might be a bit nutty. I've been meaning to add the #define and other stuff to the code for a while but I got side tracked by an IDE, Syntax highlighting and other fun stuff I'd never really thought about playing with. I've been working on #include for a while also, but I'm not going to do it until I can come up with a neat way of incorporating the line numbers and filenames properly so I can cross reference the list file and compiler errors...
Emulation is fun.. I can see the attraction of it.. I just need a few extra hours in my day [noparse]:)[/noparse] You can thank a nasty stomach wog and the resulting night of insomnia for the #define processing though :O. Imagine how much I could get done if I got a bout of H1N1!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
And core memory is edging it's way back in as MRAM. Just have to fix those reset circuits, or there will be new meaning to the phrase "My computer crashed".
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Steve
Propalyzer: Propeller PC Logic Analyzer
http://forums.parallax.com/showthread.php?p=788230
A crazy thought did occur to me some months ago: Build an ALU from 74xx. Build some registers, Accumulator, Program counter, Stack pointer or just general purpose in 74xx. Make them 16 or 32 bit. Build some interconnecting MUXes to tie these together with a memory bus interface in the architecture of your choice. So far relatively easy to do.
Now the micro-code and control unit. That's a pain, big and hard to debug. So, you guessed it, use a Prop ! It's got enough I/O pins to drive the ALU commands, MUX directions etc. You can download the micro-code easily, single step through the micro-ops etc.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
Leon
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle
Hi Leon,
Did you finish off your Prop CPLD board?
I'd like to explore the possibilities/capabilities of a Prop tied to either an FPGA or CPLD.
Seems to me an FPGA/CPLD that could handle all the I/O subsystems and memory management coupled with a Prop would make a nice and accessible dev system.
Coley
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
PropGFX - The home of the Hybrid Development System and PropGFX Lite
Sure the FPGA would be easier, and VHDL is still on my list of things to do.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
I have a couple or Prop/CPLD/SRAM/SDRAM designs in layout on a 3x4 FAB. The CPLD is Altera 7064 PLCC44 for SRAM and 7xxx PLCC84 for SDRAM. I chose PLCC since BatchPCB will produce >= 8/8 trace/space boards. The devices I fear are are too small for generic CPLD experiments though. How big a device do you think is practical for experiments?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Steve
Propalyzer: Propeller PC Logic Analyzer
http://forums.parallax.com/showthread.php?p=788230
The wirewrap is only a prototype. It will be stamped on a PCB.
I hope you will be able to see the value inherent in a design like this. Dr. Gouge has made it possible for you to do things you never thought you would be able to do with a set of Propeller chips.
Thanks for your interest.
Mark Allred
Vice President
Machine Intelligence Technologies
http://www.machineinteltech.com
Success with your new product will depend much more on support than anything else. It's easy to build a physically nice looking product. It's always a lot of work to provide documentation that helps people and sample software and I/O drivers that fit with what most people need to do or can be easily modified by and for those whose needs are different than most.
How long is a piece of string? No really I have no idea. The concept was to have a CPLD help with the access from Prop to RAM whilst at the same time providing a bunch of I/Os to replace the Prop ones lost in driving all this. In the meanwhile Cluso came up with his TriBlade external RAM system using a 10 cent helper micro controller which is going to keep me busy for a while.
Just now I'm to snowed under by "things to do" I really must resist any more shiny things. I've promised myself to one day get into FPGA/CPLD and VHDL but when I do I want the biggest FPGA I can reasonably afford. I want to swim in a sea of gates for a while[noparse]:)[/noparse] and I know there are lots of "shiny things" in the FPGA world I will want to try out.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
While memory is useful, particularly for data storage, when looking to extend program code space there's a lot more to it than just bolting some memory onto the Propeller.
I'm all for pushing boundaries, as a personal or technical challenge, to see if one can, or just for fun, but if I needed megabytes of high-speed memory I would probably be looking at something which could support that natively rather than jumping through the hoops which a Propeller demands. For $800 I could buy an Atom PC with 2GB of Ram, 160GB of hard disk and all else that a PC gives me, and still have plenty spare to buy all the Propellers I need for I/O interfacing.
Or maybe I'd just rip the innards out of a $10 router. Hmm; imagine how capable 80 would be when networked together over their inbuilt 100Mbps LAN
So here's the plan:
1) ARM or similar 32 bit CPU the cheapest we can find.
2) Some 10s of megabytes of RAM and FLASH for it.
3) It runs Linux.
4) A Prop or two with all Prop Demo board features.
5) All on the same board. No bigger than a little Eurocard.
6) Perhaps some Ethernet for the ARM
Items 1 to 3 gives the compute engine, memory, and networking.
Items 4) is there for video out and interfacing all those hardware gizmos we like to play with.
Anyone any idea how simply/cheaply this could be put together? I'm somewhat out of touch with ARMs and such but I have worked with boards like this with ATMEL chips that did not seem to be mindbendingly huge complicated projects to tackle.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
It seems to me an experimenter's FPGA board should have at least a 64K macro cell Cyclone II or equivalent.
mallred, as far as I'm concerned you can sell anything you want if you can find a market for it in anyway you like. All you have to do is find paying customers who really believe in your value proposition ... or just like collecting things [noparse]:)[/noparse]
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Steve
Propalyzer: Propeller PC Logic Analyzer
http://forums.parallax.com/showthread.php?p=788230
Before Dr. Gouge started this project, his first question was, looking at the hobby robotics market, "Where is the intelligence?"
That one question started this whole project.
Dr. Gouge has developed software that can mimic the human brain. How many software programs out there do you know of that can do this?
His genious is in taking the neurobiological model of the brain, and transcribing that into code that approximates the brain's ability to learn new information and attach to old information such that real learning occurs.
His software has been proven in several military applications. His desire is to now bring this technology to the masses, or to researchers who will continue his work when he is gone.
When he started, he looked around for a platform that would support his technology. Nothing out there was able to support it.
He asked Parallax for a debugger. There was none. So he wrote one.
He built this system out of necessity, because no platform was out there that would do this. Dr. Gouge will be building the first cognizant android the world has ever seen. So, yes, we have an agenda.
What you seem to be missing is that the Parts and Pieces that get developed along the way have far reaching applications beyond his original need for a new platform.
Who knows but what others will do with the peices Dr. Gouge creates? This is what is exciting. The possibilities are endless. We are taking the Propeller and we are doing something large and exciting with it. The pieces that get created stand on their own merits as tools and products others can use and extend.
Don't miss the forest for the trees. There is something bigger happening here than you realize.
Mark
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Steve
Propalyzer: Propeller PC Logic Analyzer
http://forums.parallax.com/showthread.php?p=788230
Go easy on the "wild claims". They've been made over and over before and they're not helpful.
The "wild claims" he has proposed are helping to protect you and me as we speak. This is proven technology in military circles.
Ok, I will give you a hint.··You are correct.· We do not know how the brain encodes information.· Dr. Gouge's technology even mimics this.· He has figured out a way to mimic the brain WITHOUT having to know how the brain encodes information.· See the Robot magazine article for more...
Post Edited (mallred) : 5/5/2009 6:31:02 PM GMT
A COG running at 80MHz executes code at 20MIPS. If it has to go to external RAM to fetch code (LMM) we're down to, what, 5MIPS. If it has to emulate some 16 bit CPU architecture (Hinted at here or on the web site) we're down to say 1MIP. If all 8 COGs are trying to run said code in parallel we are crawling.
So whatever brain emulating software there may be could for sure be run on my run of the mill dual core 3G PC with 3GIG RAM a lot cheaper than buying a truck load of Props to do it.
Also, is it really possible that some one has perfected world shattering human brain emulation software to run on a platform that does not exist yet? Surely it has taken years to perfect and is available, for show at least, on a PC.
I'm prepared to be surprised but it's going to be hard work.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Steve
Propalyzer: Propeller PC Logic Analyzer
http://forums.parallax.com/showthread.php?p=788230
I can tell you that it has been done on the PC first and is now in operation in military robotics projects. We want to create an android, and therefore needed a mobile platform to do it. Dr. Gouge chose to implement this on the Propeller.
And yes, over 40 years of Dr. Gouge's life has been pursuing this one goal, machine intelligence.
We started the project in January. Give us some time. In four months we have developed the KISS Debugger, the SRAM Memory Expansion project, the skeleton for an operating system for the propeller we are calling the KISSOS, and are currently working on voice recognition software to eliminate the keyboard. Whew! We have not even come up for a breather. Give Dr. Gouge a little credit. He has been very busy :P
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
I'm quite certain that the technology behind your company's RAM is sound, although extraordinary claims do require extraordinary proof. But, like two jigsaw pieces that don't quite fit together, your approach to the Propeller community is producing a mismatch that's resulting in some of the push-back you see here. If you'll permit me some constructive criticism:
1. Selling schematics and Propeller software just doesn't cut it with this crowd. The added value that produces sales has to come from well-crafted hardware. Kits that include a PCB (no wire-wrap stuff, please) and through-hole parts are okay; compact, completely assembled boards using SMT components are even better. What people are willing to pay for is the hardware that implements the IP. Hoarding the IP or trying to ration it by offering it for sale is not going to be a recipe for success. Basically, what this means is taking a device (which you apparently already have) and creating a product out of it. This requires a lot of extra work; but, to be taken seriously, you have to do it.
2. Your company's website seems like an attempt to create a cult of personality around Dr. Gouge, complete with photos of ID badges from his college days and prior employers. I don't quite get that, frankly. Nor will a majority of your potential customers, I'm afraid. This perception is only compounded by having you as a mouthpiece, while the mysterious Dr. Gouge remains cloistered somewhere "behind the curtain". I'm sure he has a very busy schedule, but it would really be nice to hear from "the man" himself.
I don't mean this to sound harsh or mean-spirited. On the contrary, I hope your business succeeds wildly. So I thought it would be useful for you to see how your company is being perceived. The successful businesses in this century are based upon openness, the sharing of information, and producing sound products backed by great customer service. If you want to see a perfect example of this, take a look around Parallax's website and at the things they sell, but especially at the incredible abundance stuff that can be downloaded for free — including programs from the Propeller Object Exchange.
Respectfully,
-Phil
His software will run on a 4MB expansion board. Memory represents neuronal links, and therefore, if you need your project to learn more, all you do is add memory.
We hope this approach will enable many more people to become involved.
Thank you for the input. I will definitely take it under consideration. We want this to be a successful venture, of course. If I have come across the wrong way, I apologize.
I feel we have something of value to offer the world. In four months, it is not yet polished, but we have the beginnings of something very special. My hope is to convey that in this forum, many of whom are potential customers.
The photos of Dr. Gouge on the site is because a magazine questioned his credentials, so I was asked to put those up for the world to see.
I hope to convey our desire to share this technology with the world and to create a successful business model around it. I know that service will be paramount and I pledge here and now to offer the best service possible.
Thanks again for the constructive criticism.
My Best,
Mark
So far Google has failed to turn up anything, for me, on him or his extensive cognitive research apart from what is in the machineinteltech web site.
Those pictures and such don't help, all we learn there is the he has once been a field service and customer engineer. They just deepen our curiosity....
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
On our About Us page we describe and link to some of his many patents. You might start there. If you notice, they link directly to the US Patent Office website. Also, we have posted several of his previous articles from various magazines.
He is the real deal.