THE G15 INTRAPLATE Concept introduced This entry dated February 28th, 2015. Intraplates {{{when available}}} are sold in the electronics section of Avenuege stores. {{{A few rather 'quick-typing' issues corrected and half a dozen explanatory sentences added here and there March 3rd, 2015; only those which seemed to be important for the clarity to come through. Some spelling issues remain, but we let them be.}}} Any modifications or additions, in case needed, will be added towards the completion of this page, which is at: avenuege.com/intraplates by Aristo Tacoma Contact info: completion of this paper THE INTRAPLATE A novel electronics concept, part of the pathway from the G15 software package at norskesites.org/fic3 to the realisation of a computer running the G15 directly as O.S. design, suitable for a range of also research oriented first-hand electronics work Illustrations composed via GEM in G15; note that the notion of 'modifiable' refers to a computer made of such units Location of this whitepaper: http://www.avenuege.com/intraplates This paper concerns hardware which is oriented towards educational electronics, and also for production of the computer which runs the G15 Yoga6dorg instruction set directly. For free download of the G15 Practical Virtual Implementation for Linux and DOS pls consult: norskesites.org/fic3 Slogan: the intraplate is a golden middle-way between the traditional transistor and the microchip concept, allowing human first-hand electronics also when complex electronics requiring some compression is required, without relaying on eg nanoscale chip wafer foundries at all Keyword: "intraplate" is hereby defined as an extremely compact way of making available to the human electronics person human-maintainable massive bundles of transistors (or transistor-like components), capacitors, resistors and such elementary electronics components available by something scaled between the traditional transistor concept, and that of the microchip; and with wires, rather than pins; and so that whole computers can be designed from scratch without reliance on extremely costly microchip foundries; there are also numerous educational electronics applications of our new concept of intraplates in electronics. BACKGROUND: HISTORICAL CONSTRUCTION METHODS OF COMPUTERS ======================================================== The computer concept arose in part due to speculations over the foundations of pure logic and number theory by the works of Alan Turing, directly in the wake of Kurt Goedel's famous 2nd incompleteness theorem. Goedel, as is well known, managed to utilise the systematization over logic and number theory brought about by Bertrand Russell and Alfred North Whitehead in their Principia Mathematica to show that self-reference is implied by their typed set theory -- a self-reference that can only be avoided by assuming that their set theory is incomplete. (The alternative would be inconsistency.) Alan Turing evoked the notion of a person carrying out certain procedural rules of computation -- a so-called "Computer" -- in order to erect a thought experiment connected to the meta-method that Kurt Goedel employed in his article from around 1930. For, as Turing pointed out, it seemed as though Goedel, by what Goedel called a form of "meta-reasoning", was able to go beyond the limitations of the rule-based axiomatic system he was investigating. Could then not this approach of 'going beyond' become itself the seed of an entirely new way of reaching a sought completeness, which is also free from self- contradictions? While Turing didn't achieve this result, but in fact showed that Goedel's type of incompleteness shows up again and again at meta-levels, the immensely fruitful concept of the computer machine was thereby born. It took, however, the Second World War to provide Turing with full finances to put together, by the electronics means at the time -- involving electron vacuum tubes and a set of mechanical switches and such -- a first simplistic general computer; however Turing had conceptualised much before this, and still more after this, and reached notions covering essentially not only all of algorithmic programming in general, but also such as the type of (still algorithmic) programming involved in 'pattern matching' programs many decades after his death. A great many contributions from other people are of course necessary to bring in, if one wants to give a full account of the history of digital computing; there are also interesting pieces of machinery and of thinking along these lines by other people long before Turing's work. One can also then bring in discussions as to whether Turing had somewhat overheated assumptions as to how far a computer could go, -- in other words, that he perhaps didn't fully take in over himself how fully the "goedelian incompleteness" (as we can call it) would still imply that something breaking with all algorithmic approaches, and thus also breaking with the whole digital computer concept, would be necessary to more fully understand what he and a number of other philosophers have tentatively called "intuition". This is outside of the scope of this article, however, which is concerned solely with the physical means by which computers are made. The role of the electron tubes in the earliest computers is, as is well known, vaguely akin to that of a composite valve of electricity. A gentle flow of electricity into a tube can set off a strong current of electricity; and by suitable modifications of the design, it can be set up so that at least two such mild flows of similar current must go into the tube for one strong flow to go through it -- or that at least one of two such flows into the tube is adequate to start the strong current. By a little modification of the setup of some tubes, one can get the arrangement that only when a mild current is NOT given to a tube, then a strong current will indeed flow through it. We see here parallels to the boolean logic operators of AND, OR, and NOT, and the related operator of XOR. It has been shown that given some means of temporary storing a voltage value as either 'high' or 'low', -- parallel to the notion of 1 and 0, or "one bit" -- combined with not much more than a big bundle of such AND, OR and NOTs, one can construct the addition operator for numbers represented in the socalled binary decimal system. While this is simple in theory, the fact is that the quantity of such gates grow almost exponentially for each additional bit required, and even with as many as 16 bits, one still gets numbers not higher than about 65536, or signed numbers up to about 32767. To achieve a more psychologically convenient range of up to about plus minus two billion, one would need as much as twice that many bits again, or 32 bits. And still more boolean gates are needed for more operations, like multiplication; and to actually construct all the operations necessary for a general computer one would have to have a whole range of additional operations, each with a more or less great requirement of logical gates. In praxis, this means that with a starting-point where something like three-finger- sized electron vacuum tubes are used for the gates, even merely an 8-bit computer would easily require hundreds of square meters and corresponding enormous amounts of electricity. As soon as germanium, and later silicium, and some other substances with some metallic features, called "semiconductors" (half-conductors) were found (with suitable microscopic additions of such as antimony, boron and more to highly purified semiconductors) to be able to act as logical gates for much smaller voltages, and without heating required, these began to replace electron tubes not only in radios and amplifiers and such, but also in computers. But the semiconductors had new challenges connected to them, and one was brittleness: they easily overheated, in which case their fine crystalline quantum field derived gate-like functionality would no longer be available. And even with these much smaller entities, something like a 'home computer' or a 'personal computer' was entirely unrealistic: for while such a computer ought to have in an extremely minimal sense at least 16 bit, but more preferrably a full 32 bit range, capable of representing huge catalogues of both books and images and also to advanced permutations within its address range, even a relatively weak 16-bit computer of the transistor (or transistor-like) type still took up considerably more size than a living-room. And then only with powerful fans enabled, and most rudimentary input- output devices such as paper cards with holes and textual only paper-printed output, at least initially. The electron tube was, in the phase, effectively replaced by the transistor, and related forms of semiconductors, with much the same type of idea of its functionality. Each transistor (or related concept, such as thyristor, MOSfet and such), just as each tube, requires a couple of very simple components to modulate its operation -- notably the resistor (which we in our works often call a 'modulator', for it modulates voltage), a capacitor (which is little but two metallic plates near one another without touching), and the coil. Incidentally, the same components are also used in radios, and radios, just as computers, need oscillations in order to drive their processes. Still, with this type of transistor computer, a great deal of developments took place in universities and other research institutes in the 1960s. In the 1970s, the quest took a leap, at least commercially, by the successful production of microscopic-sized transistor-like components, wired up together with equally microscopic-sized capacitors, resistors and coils. This step, a result of contributions by many as all the other steps, was successfully (in a commercial sense) fulfilled first by the company Texas Instruments, a Texas, USA company. Already in the first years of the 1980s, the IBM company launched its first Personal Computer, and established a set of ways of doing hardware which became embedded, and elaborated upon by many, in the following decades of "PC" productions. Throughout all these years, the notion of microchips was dominant, and although this meant that, efficiently, all design of the core computer was taken entirely away from engineers doing their own soldering and fitting together of individual components like transistors and capacitors and such, it became commercial viable and technically possible to make more and more bits and more and more computer memory compressed into less and less space, and at a price which appealed to the many. In this process, a number of companies, including such as Intel and AMD, arose, which produced whole computers integrated into chips at their 'wafer foundries'. However the extremely stringent conditions under which such microchips are developed meant that all essential computer development could only happen at a few extremely costly places on the planet. This is in contrast to transistors, which to a fair extent can be made by means of such as a high school chemistry lab when equipped with not too complicated extra equipment and not too hard-to-obtain substances. While people of course were at liberty to buy these microchips and fit some components around them and in that sense 'put together their own computer', the tremendous complexity of a large part of the CPU- oriented microchips (with the exception of those consciously made with a 'Reduced Instruction Set' notion), and the complexity, then, with many of them, also of the underlaying 'machine language', meant that, in some sense, by the advent of microchips, the bottom had fallen out of the enterprise of hobby electronics. Hobby electronics became now much a matter of wiring up that which is predesigned to work in various ways -- whether as full computer microchips, or more general-purpose less complicated chips with a bundle of microscopic components in them, and 'pins' sticking out to which one can fasten one's wires to other such components, as well as to traditional components. Those who wanted to have more complex functionality coming straight from a chip could buy kits where certain sections of a microchip is capable of holding some programs, so as to burn in some programs into the chip itself -- but still, this type of so-called "ASIC" or Application Specific Integrated Chip design relied fully on the outputs of the industrial wafer foundries plants, of which there could only be a few in the world, due to the immense technical complexities involved in designing something involving such microscopic sizes. NEED FOR ALTERNATIVE TO THE MICROCHIP CONCEPT ============================================= When one reads in literature connected to the research of the microchip foundries, this research is typically oriented towards squeezing yet more components into even less space, or in a way that requires somewhat less cooling or a material which involves greater reliability. There's also research on other types of comptuters, where the notion of the digital boolean gate is imagined to be just one of several, and where it is sought -- but perhaps without any single convincing result so far -- to exploit and manipulate more elements of that which is pointed to by quantum theory than those features already called on by the notion of semiconductors (which is wholly a quantum concept). One can imagine the commercial agendas of the few privileged places on the planet competing with each other, in a little club, to make microchips. However, for most people with a technical interest, it is of great importance to be aware that an aspect of the situation remains untouched by the endavours of these giant, rich, super-specialised factories, an aspect that we have hinted at in the previous section. And that is the fact that only up to the transistor computer, was there a capacity for the individual technical expert, working on a local computer installation, to have a direct contact with the actual design of the computer and even to be able to affect it, change it, and repair it when need be. For only as long as the components were of a kind that could be, at least in principle, be handled by people, could the design of the computers be said to be in any true electronics sense to be open and transparent -- at least relative to the hobby electronics person. Now one might argue that the hobby electronics interested individual can still read the design description of the chip (insofar as it is provided in detail by one of the chip foundries), and such an individual is still free to get first-hand experiments e.g. by tinning together such as a radio. What would be the need for individuals to have more of a direct say in the underlaying computer structure when this is so inexpensive and easy to acquire by means of using any of the inexpensive premade chips? To understand why this type of response isn't satisfactory in the long run one must take a step back and ask for the greater evolution of the consciousness of humanity in the technological era, and ask what type of self-education process humanity is in, and OUGHT to be in, relative to what directions technology ought to take. The microchip production is eminently suited to create a power monopoly on behalf of those very few individuals in charge of microchip foundries. The larger currents of self-education in humanity involves also such questions as how does science affect the understanding humanity has of itself and nature, and cosmos as a whole? And how can one make science come alive during the education process of the also very young? One can see here the value of the much-quoted statement by the hellene philosopher Aristotle, dating to about half a millenium before Christ: "A small error in the beginning grows to a large error eventually." And in the self-education process of humanity, we must certainly regard the shutting away of the process to have a first-hand connection, directly, to the design of one of the most-used technological instruments in all humanity, as an error which, however small it may appear in the beginning, can become a grave error if it is allowed to proceed in unfettered ways without any challenge. To frame the question more positively, isn't there something that can be done, in order that the core computer Central Processing Unit, or CPU design, can be made more transparent, open, and possible to both maintain and change for individuals? Must there be a complete goodbye-saying to soldering more or less individual transistors and related components in an advanced age of electronics where the making and modification of computers and computer-related piece of electronics are essential points? And indeed, if there is a way which doesn't have the challenges of the early transistor computers, nor the challenges of the microchip foundry factory based computers of the just-mentioned sort, we should go on this way. And, the G15 PC concept, first made in software as a virtual CPU run on top of existing commercially available computers, has led us to construct a solution, a pathway, and we call this the G15 INTRAPLATE. It is here presented for the first time in detail. It is a novel concept not in terms of what it achieves technically, but only in HOW it achieves it, in its 'humane' or 'hobby electronics' radiance, so to speak. THE INTRAPLATE CONCEPT AS STEPPING-STONE TO G15 HARDWARE ======================================================== This concept paper, with this and the next section, and the two former sections, is written while the software of the G15 is existing, with a virtual CPU emulation on top of Linux and the G15 Yoga6dorg assembly performing well also in a higher-level script language with some inspirations from Forth and many other languages, which we call PMN. It is written before we have done serious hardware work. This concept paper is to explain the work process, and set forth the concept -- the Intraplate concept. This word has some uses in geology, connected to subterrenian plates and their movements during earthquakes and such. In our use, which is entirely more peaceful and less dramatic, we use the word 'intra' simply in the sense 'existing within boundaries, a connection within a domain', and 'plate' in the sense of an electronics element with some complexity, somewhat like a microchip and yet not a microchip. The idea here of 'intra' is threefold: each plate have some inner connections, but not so many as to exclude further design in hobby lab usage; several plates are normally required for any advanced design, and so they are connected; and thirdly, while this is admittedly a little more complicated than the making of transistors, it is not much more complicated -- this is, then, an INcorporating of TRAnsistors (IN-TRA) (by which we here mean not just the classical three-pin NPN or PNP transitor, but also related semiconductor types), among other suitable components in a way which is not microscopic, merely compressed by precision work. {In contrast to the transistors, though, intraplates have a compression and precision factor which obviously call on some minute precision-work most suitable for elementary kinds of factory-like robots, with robotic movement possibilities, robotic arms, and a simple camera.} While, with the microchips, the aim is to put as much electronics components in an as little piece as possible, the concept of the intraplate is to put a group of electronics components together in a way that allows the construction of such as digital computer CPUs and related components out of them. The aim with the intraplate is to put enough components that the quantity of intraplates required to make up a whole digital compute with both CPU and RAM and such as monitor, keyboard and mouse controllers constructed out of them isn't all that great, while each intraplate isn't so compressed that first-hand electronics work cannot be done with it. The intraplate concept overlaps with the functionality of a certain type of general-purpose microchips, which do not by themselves have anything like a full CPU on them, but which, when wired up with many other similar chips, can be used to make up a CPU somehow. For there are, indeed, very many various microchips in commercial existence today, in addition to dedicated full CPU microchips. But the intraplate has a design objective which is different: it aims to make it unnecessary altogheter to use a CPU microchip, and it also aims to make it as easy as possible to construct a CPU and the rest of a computer by means of itself. The intraplate, then, first and foremost, is a set of condensed transistors, capacitors, resistors and such, with a minimum of pre-designed-ness about them, and made in a way that doesn't require extremely complicated factories, and which has a format that allows hobby electronics work in a way which is genuinely open -- not merely a question of wiring predesigned circuits. We are speaking of a whole different approach to such as digital electronics construction -- and part of an advanced level of what we elsewhere have called "Elsketch", or learning by 'sketching' electronics. The intraplate is in general larger than the more or less nail-sized microchips commonly in existence. The quanity of connectors to each intraplate is by necessity huge, in order that it is possible to regard each intraplate as also a first-hand hobby electronics concept where what has been acquired is simply a set of very small components rather than a premade wiring. One of the distinct challenges for hobby, or more generally, research-oriented and first-hand electronics in dealing with nail-sized chips surrounded by myriad tiny 'pins', many of which stand very close to one another and can easily break, is to be able to not only design, but also redesign connections involving them. In the standard proposed intraplate concept for G15, each intraplate does not only have a more robust size, but it comes with labelled bundles of individually marked plastic- insulated copper wires. In order to construct computer elements out of a group of intraplates, one doesn't need to engage in high-level precision board circuit design of a type that cannot be easily altered if the design isn't working all that well. Instead, one can connect, disconnect and reconnect wires individually by patient precision work with the same type of tinning instruments and such as used in the 1970s hobby electronics laboratories. But the operation will be able to proceed into real computer design, for the intraplates involves thousands and thousands of condensed transistors with associated components, and a similar number of wires, made so that it makes economical sense, and sense also in terms of sizes and power supplies required, for individuals to do electronics work with them. The intraplates that we propose are that of a whole conceptual range, some more adequate to make computer CPU, some more for its RAM memory, some more for associated technical input/output tasks, some more for network connectivity, some more for such as disk control, etc. CHALLENGES THAT CAN BE MET WITH INTRAPLATES IN COMPUTING ======================================================== We shall mention here some of the challenges involved with engaging intraplates, also so as to show that the component of cool realism is required -- there aren't only advantages over microchips once we start to use intraplates to make computers of them. There is a set of meetable challenges, so to speak. One of them concerns the degree to which complexity at the CPU level can be tolerated. The G15 PC has been designed by means of around 250 instructions, most of them utterly simple, although some of them presuming some specialised electronics e.g. for mouse pointer showing or keyboard reading and buffering, or for pixel-refresh, as well as some loop operators. But this bundle of instructions is so that it is after all enormously simple compared to the level that with the microchip industry became typical in widely available computers -- yet without there being any clear reason in favour of this complexity at the core level. The core level ought to be simple in order for it to be stable. So this challenge has been met, and -- as any peruse of the G15 y6 package will show, successfully. Another challenge is that any distance and any use of such as copper wires when things are oscillating, as they are, at the level of many megahertzes, means that there are sensitivities involved connected to radio waves being generated and transferred between the elements in the process. This challenge increases the higher the quantity of megahertz increases. This challenge must be met in a twofold way, at least: one is that the G15 PC is designed so as to accomodate a moderate, and not overly intense, degree of megahertz, as a design component which is inherent even in the language. (This has been achieved.) The other way to address this is physically, by means of powerful radio wave insulation elements e.g. made of gold inserted into construction at suitable points, and with care taken that components most likely to produce and be wired into fast and relatively strong oscillations are positioned smartly near one another. The intraplates will need cooling, and presumably, while there are certain advantages when it comes to this aspect in the increased size of the computing machinery when it's made of intraplates rather the microchips, the role of active cooling is likely to be such, in most intraplate computers, that a considerable space of the computer must be set aside solely to ensure great cooling. The size of the computer, and its weight, is obviously itself a challenge. We can only say that if it can as easily as a fridge fit into a home, and with all the advantages relative to self-education, home computing, and being part of an optimistic first-hand spirit relative to what technology can mean for the human race, then this will be an acceptable size. Finally, even with all the aims going into simplifying the design of the core components of a real 32-bit Avenuege G15 PC with several dozens of megabytes of RAM in it and an assortment of natural devices connected to it, including a 1024*768 greentone monitor, a full QWERTY Ascii 7-bit keyboard, mouse, disk, net connection to other such PCs, and means for backups, we are talking of such a number of intraplates and such a vast number of wires coming from them that the whole PC, in this case, is a daunting task for any individual to put together even given many seasons of full-time work dedicated to just this. What instead has be done is to say that when the work must be oriented towards the full Avenuege PC construction, in order to make several such with fair ease, we need some automated robotic arms and such to do the main bulk of the connections. This is however compatible with the idea that it is still a product that, even at this level of complexity, has a PRINCIPLE of openness to first-hand electronics work maintained throughout itself. The individual transistors, which can be made in any relatively well-equipped laboratory given very pure samples of the right conductors, are almost, but not quite, products coming straight from nature. The notion of condensing them, while requiring some extra machinery, is not dependent on superspecialised plants utilising electron microscopes in their design-- it is simply a more precise, and somewhat smaller process of putting together transistors -- along a continuum. So the principle of first-handedness, of openness towards hobby electronics, and of a preference to only use components that are more obvious in their composition, and nearer that which can be made straight from Nature, is maintained in our intraplate concept. ====================================================== ====================================================== Contact info for author of the above, Stein Henning Reusch Braten {son of professor Stein Braten and his wife Else Reusch Braten}, with pen names derived from family history names including Maria von Weber, and artist name includes Aristo Tacoma: h-reusch@frisurf.no Yoga4d von Reusch Gamemakers Holmenv 68 N-0376 Oslo Yoga4d von Reusch Gamemakers is a private company registered in Oslo, Norway, wholly owned by Stein Henning Reusch Braten. It is he and his private company, which also owns the Avenuege stores, which has the copyright to the Intraplates concept in electronics, the Avenuege G15 PC concept, and to the range of terms associated with the G15 Yoga6dorg PMN product, also as informal trademarks enforced by use. There are however generous licenses associated with the great quantity of source released together with the G15 platform, inviting both research and commerce, as long as acknowledgements and such are intact.