Visions of the Future of Computing Professor Peter Excell Professor of Communications / Athro Cyfathrebu Glyndwr University / Prifysgol Glyndŵr
Computing: Where Have We Come From? Computers, Communications and Broadcasting are CONVERGING Computers started over 60 years ago –Initially seen as mathematical But - Alternative view: computer as a communications device (c.f. Colossus code-breaker)
Where Have We Come From? 60s/70s: networking and Arpanet Internet 60s: Telephone networks adopted digital (PCM) systems Telephone Exchanges hence become a form of computer
Parallel and Vector Computers A Big Iron view of the way forward SISD: single instruction, single data SIMD: single instruction, multiple data (MISD: multiple instruction, single data) MIMD: multiple instruction, multiple data
Parallel and Vector Computers Flops: floating-point operations per second MegaFlops – 10 6 GigaFlops – 10 9 TeraFlops – 10 12 PetaFlops – 10 15 Exaflops - 10 18
Parallel and Vector Computers Success in Vector Processing: Seymour Cray Cray-1 at London Science Museum (P. Excell)
Parallel and Vector Computers Crays philosophy: build maximum-speed serial processors first (vector processors), then parallel them if necessary. Crays career: CDC 6600, 7600, Cray 1 - vector processors for SIMD (not parallel as such) Then cautious parallelism: Cray 2, X-MP, Y-MP Then more parallelism, using torus structure: Cray T3D, T3E Cray was much more successful than his rivals, but he relied on Cold-War defence funding. Japanese fifth-generation project: machines following Crays philosophy from Fujitsu, Hitachi and NEC Problems in parallelisation (or vectorisation) of problems and software: some think it should be transparent, but that often seems to fail.
Parallel and Vector Computers Parallel wreckage in the dustbin of history: Floating Point Systems Corp. Kendall Square Research Corp. The Connection Machine (Thinking Machines Corp.) Convex Computer Corp. Inmos Transputer Meiko Computing Surface CDC Cyber 205 computer (Cray-1 rival) eta-10 (CDCs last fling) Parsytech GmbH Cray-3 and 4 computers (planned but never built) occam parallel programming language Dataflow computer architecture
Parallel and Vector Computers Structures: Parallel Virtual Machine (PVM) Message passing interface (MPI) Beowulf systems Grid Multicore Cloud
Parallel and Vector Computers What about Quantum Computing? Seems to be a sort of super-parallel machine No clear view of when will become viable Could be driven by intelligence needs
Apple 1 and Altair Apple 1 and Altair - National Museum of American History, Washington
Where Have We Come From? Computer communications: at first for serious uses Early PC uses also serious PC business almost died, then Cold War ended Mass market had to be appealed to Consumer applications threw up worthwhile new challenges Moores Law: a major influence on strategy
The Home Computer was evolving incrementally (e.g. Apple II), then IBM stormed in, rushing to get into the PC business. 3 dreadful business errors meant IBM threw away its hardware and software: –Compaq hacked BIOS (legally), because it wasnt protected properly: this led to the burgeoning of PC clones –The best OS company (CP/M) was scared off by non- disclosure agreements: the OS business was then given to Microsoft on a much weaker basis –Microsoft was assigned rights to all software The Rise of the PC: An Extraordinary Shambles
Further: The GUI and mouse concepts were developed at Xerox PARC – not seen as of value to Xerox Essentially copied by Apple Then copied by Microsoft Other business failures: BBC Microcomputer, Acorn, Sinclair, Amstrad, Commodore, Tandy.. The Rise of the PC: An Extraordinary Shambles
Where Have We Come From? Communications: Optical fibres huge capacity packetised messages Internet Crucially dependent on a computer at each end of link GSM revolution (digital mobile phones) Highly complex - but cheap processors - mass market justifies huge investment, leading to commoditisation Upgrade path mapped: 2.5G – 3G – 3.5G – 4G 5G – millimetre-wave??
Where Have We Come From? Mobile is established part of everyday life worldwide No. of subscribers over 4 billion –i.e. about 3 people in 5 in the World –Will be 1 in 1 soon (already reached in Europe) (c.f. PCs – market saturated at a much lower level? Is this market dominated by relatively affluent males??)
Where Are We Going? Mobile phone: has more processing power than recent-past PCs Mobile has increasing Internet ability All it needs to replace the PC is storage and better HCI (human-computer interface): –Miniature hard disks already available (e.g. in iPod) –Input: e.g. voice recognition, handwriting recognition, folding/miniaturised keyboards – 3.5G laptops –Output: larger video displays in phones, unfurlable screens, head-up display devices
Where Are We Going? iPhone: amazing triumph of coolness plus usability Blackberrys response: Storm/Thunder Google Android – new player OVI – Nokias answer
iPhone problem (until iPhone 4) – or is it? Future Concepts
Additional Functionalities Bluetooth - Earpieces; Handsfree; Netbooks… UWB (Bluetooth 2) – much greater bandwidth RFID - major opportunity: mobile wallet; rewriting tags… Glyndŵr RFID stolen tax disc checker
Additional Functionalities GPS E.g. Branded meeting places project. Edinburgh & Glyndŵr Univs. Tagged memory places near Edinburgh Univ
Where Are We Going? Paper-like unfurlable concepts - ~A5-size output device
Where Are We Going? Alternative input devices - Haptics Alternative output device - Eyepiece display Gaming/video devices: Good, but block vision
Where Are We Going? Military in the vanguard: Full head-up display
Where Are We Going? Watch computer/ mobiles: – Stalled – why?
Where Are We Going? Raft of ambitious roll-out plans –3.5G, 3G-LTE, 4G, 5G - evolved from GSM SMS MMS; WAP WAP 2 Mobile Internet; email; 3G with video bandwidth (why not used much?); 4G will have more bandwidth than needed for video…. WLAN (e.g. WiFi or IEEE 802.11a/b/g/n) a rival technology – was thought might merge with 3G to form basis of 4G – now a sideshow Mobile WMAN (e.g. WiMAX or 802.16e/802.20) – another rival for 4G – beaten by LTE PANs: Bluetooth (802.15.1); UWB (802.15.3); ZigBee (802.15.4) Mobile TV (DVB-H; DMB)
Where Are We Going? Can we get a clearer view? Can we avoid the mistakes of the past?
Technological Forecasting A necessarily imprecise management tool that attempts to predict the evolution of technology in the near-term future.
Why is it important? All technology-based companies, and most individuals, have to form a view about the near-term evolution of technology, to decide on investments, purchases, life plans and career paths. We are living through a particularly exciting time in the evolution of technology: Computers; Mobile phones; (insert your favourite here)
The Kurzweil Singularity Peter Cochrane (ex BT futurologist): In my estimation it will go something like this: 2006 internet ~ 1 human brain 2012 internet ~ 1,000 human brains 2018 internet ~ 1,000,000 human brains 2024 internet ~ 1,000,000,000 human brains 2034 internet ~ 1,000,000,000,000 human brains Already, we can read peoples minds, crudely, using MEG (Magnetoencephalography): by 2100 it is reasonable to expect that this will have enough resolution to enable the complete brain contents to be downloaded
Futurology: an Aid to Insight Millimetre Timeline focuses thoughts: 1mm 1 year 1 metre 1000 years 1km 1 million years 1000 km 1 billion years So: The Renaissance ~40 cm Industrial Revolution ~20 cm Computing ~6.5 cm - The blink of an eye…
Digital Examples The PC, contrary to the expectations of only a few years ago, has become a widely used device and the dominant form of computer. And yet it shows signs of having reached a plateau. Meanwhile, the mobile phone has become even more widespread around the world and appears likely to acquire all of functionality of the PC within a short time.
Personal Perspectives Several technologies have effectively died out over recent decades: can we predict when this will happen, to avoid damage to our careers? Patricroft steam locomotive depot, 1968 Radio Valve
Personal Perspectives Recent deaths: CRT screens Incandescent light bulbs Analogue TV Film cameras Floppy disks Vinyl records Tape? 1G mobile
Personal Perspectives Several promising innovations have also failed to take off in the way that was hoped: can we predict this better? Gas turbine vehicles – Hovercraft – Concorde – Electric vehicles – Betamax – Superconductivity – Transputers – Sinclair QL – Sinclair C5 – Minidisks… What should we invest our time and money in? There is thus a need to have a vision of where present and future technologies are heading.
Personal Perspectives C.f. The remarkable failures of forward vision in the mobile industry: The unexpected success of SMS The marketing debacle of WAP-1 The extraordinarily low uptake of video telephony The unexpected success of the iPhone
Great Blunders in Forecasting I think there is a world market for maybe five computers. Thomas Watson, chairman of IBM, 1943 Computers in the future may weigh no more than 1.5 tons. Popular Mechanics, forecasting the relentless march of science, 1949 It would appear that we have reached the limits of what it is possible to achieve with computer technology, although one should be careful with such statements, as they tend to sound pretty silly in 5 years. John von Neumann, computing pioneer, 1949
Great Blunders in Forecasting British government calculation, early 1950s, to establish the size of the market for computers: Computers do arithmetic, so estimate the total number of arithmetic calculations being done by human beings (principally financial clerks) and calculate how many contemporary computers were required to replace this work… Answer: 4 (but add a 5th one in Scotland for political reasons)
Great Blunders in Forecasting There is no reason anyone would want a computer in their home. Ken Olson, president and founder of Digital Equipment Corp., 1977 640K ought to be enough for anybody. Bill Gates, 1981
Great Blunders in Forecasting By the mid-1980s the home computer boom appeared to be nothing more than a short lived and, for some computer manufacturers, expensive fad. … As a result a device that was initially heralded as the forerunner of a new technological era was a spectacular failure that threatened to bankrupt the firms that had invested billions of dollars in its development. From "The Evolution of Technology" by George Basalla (Department of History, University of Delaware), Cambridge University Press, 1988 (reprinted 1995)
The S-Curve Horizontal scale is for illustration only: measures of technological parameters and evolution durations vary widely. Tech. Parameter = 1 represents the physical or natural limit of what is possible
The Industry/Technology Life-Cycle Scales are for illustration only. Key problem: defining the boundary of a particular technology
Triggers for the Decay Phase Michael Duffys concept of Design Impasse: improvements are physically possible, but effort is better directed elsewhere (Duffy, M.C. Technomorphology and the Stephenson traction system. Trans. Newcomen Society, Vol. 54 (1982-83) pp. 55-78) E.g. The Sailing Ship Effect; Steam engines
Moores Law Gordon Moore, of Intel, made a prediction about the rate of increase of processing power (or intelligence) of microprocessors. He said that the number of transistor gates on a silicon chip would double every 18 months. He made his prediction around 1965 and, amazingly, it has held true ever since. This is known as Moores Law. Not a law of physics: it is partly a self-fulfilling prophecy, because Intels shareholders would think the company had a problem if it could not maintain this rate of progress, but would suspect it was spending too much on R&D if it exceeded it.
Moores Law This has led to Commoditisation of computing, e.g.: Ubiquitous/ Speckled computing RFID tags Soon to be everywhere, not just a replacement for barcodes, they have intelligence and rewritable memory IPv6 addresses Will enable huge numbers of artefacts to have an Internet address
Delphi Method for Futurology Delphi Method: iterate expert opinions The simple artifice of keeping the experts apart and keeping their responses anonymous frees them to think radically, creatively and independently, and this gives the most useful result from the exercise.
Are Computers Different? Computer and software technologies appear to be qualitatively different. Because they possess a degree of 'intelligence', they have more ability to 'reinvent themselves' and hence break the life-cycle pattern of the past. E.g. the PC is almost an entirely different product from 1 st -generation computers A PC running Word is effectively a different device from one running a game
Are Computers Different? This raises questions about the definition of a Technology in the Technology life cycle - Will computers as such become obsolete? - Doesnt look likely?
SystemIntro Year Bandwidth 2G (GSM)19919 kbps 2.5G (GPRS)200128 kbps 3G2003384 kbps 4G (LTE)2010100 Mbps Moores Law and mobile bandwidth
Bandwidth A measure of the rate at which information is transferred Analogue version: frequency bands –Voice needs about 5kHz –TV needs about 5MHz –So, a picture is worth a thousand words
Bandwidth 802.11a/b/g WiFi: 11/22/50Mbps 802.11n – aiming for 500Mbps guaranteed 802.16 WiMAX: 70Mbps Pace of change is faster than Moores Law All brilliant technology, but what will it be used for?
Human Communication Bandwidths Human Input –Very high quality video (~100Mbps??) –Good quality audio (~100kbps??) –Other sensations – hard to quantify Human Output –Low quality audio (~20kbps?) –Typing (~1kbps?) –Gestures and body language – hard to quantify –Does deaf sign language give us some clues?
Human Communication Bandwidths So the human beings communications abilities are very asymmetrical (by ~5000 times) Compare this with ADSL (popular broadband technology): A stands for asymmetric: the system uses 4 times the capacity for downlink traffic (i.e. to the consumer) as uplink (from the consumer) Is this a good thing? Should we facilitate people being more creative??
Future Concepts Clear logic of convergence is that a high-end future mobile will have –processing and storage capabilities similar to a good contemporary PC and –communications ability in Megabits/sec (WLAN/ WMAN/ 4G) Adding convenient and not-too-conspicuous human-computer interface means device will have to be distributed around the body –in other words it will become the same as the wearable computer but with a wide-band radio modem
Future Concepts Steve Mann, University of Toronto - Evolution from geek to cool
Future Concepts MIT Media Lab – espousing the integrated-garment approach (How do you clean it?)
Future Concepts This vision could lead on to what may be called the 'Borg paradigm', drawing on the concept explored in "Star Trek -It doesnt have to be like that! P.S. If theyre so smart, why didnt they think of Bluetooth?)
Xybernaut POMA (Hitachi WIA – Wearable Internet Appliance) - An attempt to get cooler Future Concepts
Icuiti headset -With laptop in backpack, and handheld trackball -But with OQO mini tablet computer, it fits in a pocket Future Concepts
Augmented reality headset for mobile Future Concepts
Karlsruhe University lip-reading system Future Concepts
More radical ideas need to be considered for the way in which it would be used Cabling objectionable – use wireless (e.g. Bluetooth) – leads to BAN/PAN concept (Body/Personal Area Network) But: charging or other power supply a problem –Energy harvesting needed Wearable components: –CPU, possibly separate storage unit, –combined retinal projection unit, earpiece and microphone, –wrist-mounted keypad/mouse unit, –radio transceiver (on belt or in shoe to provide power)
Future Concepts Think laterally! Generator (later transceiver) in the shoe
How would such a system be used? Video phone communications on the move? Certainly, but not all of the time. E-mail, with optional added pictures or non-real- time video? Yes, this is likely to be a major source of traffic. Web browsing? Yes, but not much of the time. Employer information? Very possibly, but not much outside working hours. Immersive gaming? Possibly, but not for the majority and not in working hours. Is this all? Mobile TV? Rubbished until recently; now might make it?
How would such a system be used? Need to break from the Microsoft Windows paradigm: Desktop Folders Files
HAPPI: One Vision of the Future Personal advisor: a vision from 1980 (IEEE Spectrum) A cartoon, but the idea has merit
HAPPI: One Vision of the Future HAPPI: Human-as-peripheral paradigm initiative Hence human user will become an appendage attached to global intelligent network Standard caricatures depict a frightening and dehumanising experience: justified? Consider user with mental handicap – computer/network acts as helper/advisor
HAPPI: One Vision of the Future Should such aids be limited to the handicapped? –If network has potential to be more intelligent and knowledgeable than even the wisest human being, surely it has potential to improve life of any person integrated with it? Deep questions arise about nature of intelligence and value and purpose of human life
HAPPI: One Vision of the Future This is a good thing: computers and automation have already freed human beings from doing many monotonous jobs, –so the wearable computer/communications system has potential to free the human from the great majority of tedious and non-creative activity –human would be freed to concentrate on faculties that cannot be replicated in a computer –the human would become akin to a special-purpose peripheral, with unique abilities, attached to the computer system
HAPPI: One Vision of the Future This vision could be realisable within about two decades: –researchers and industry need to explore such possibilities (and consider alternative visions) –preparations need to be made, especially in development of the relevant software/content –potential is challenging, frightening and exciting. It seems likely to change human life more than any other technology –Even those who object to it will be unlikely to be able to prevent its evolution
The Emotional Wardrobe: another vision of the future Project (with Glyndŵr staff) led by garment design researchers at Central St. Martins School of Art & Design, London (CSM). Investigating technology-enabled garment design CSMs specialism is communication of emotions, by sensing skin potentials etc and communicating this as warmth, touch (hugs) and displays –This represents a new human output channel
The Emotional Wardrobe: another vision of the future
Immediate tasks Mobile Content Development Expected to be a big business (4+ billion customers) Needs creative ideas We are developing teaching of this New scenarios are currently mainly research, but also some student projects
Mobile Content Hello Kitty: culturally significant Japanese 2D barcode sign
Aesthetic animated image: From midlet.org Mobile Art
Mobile Content The problem of screen formats: Old Nokia 9210 format
Mobile vs PC Mobile devices have bigger economies-of-scale advantages They digitally enfranchise more people – in both developed and developing worlds They facilitate spontaneity – major creative aid Interest in fixed phones is fading: fixed PCs could be next But much more needs to be done to make the mobile device a facilitator of enriched life.
So what about traditional computing? Computer power trickles down to smaller and cheaper units The screen+keyboard model wont last for long Experience with cutting-edge computers helps us to envision what will be commonplace tomorrow HPC Wales project coming….
Future Ubiquitous Computing Integrated Product Design Approach: 1.Engineering design (of the internal electronics) to function reliably and at an economic cost 2.Physical product design (aesthetics, ergonomics) 3.Software design, to achieve a useful/desirable purpose reliably and at an economic cost 4.Human interaction design: aesthetics, meaningfulness, engagement… 5.The business model (its no good making a loss!)
Conclusions The future will be very exciting for computer-based technology. Our wilder ideas are more likely to come true than conservative visions. Moving from the big-government leadership of technology in the Cold War era, to the consumer-led situation today has been far better for progress than we imagined. Liberating all the people in the World to do more creative things is a wonderful opportunity that computer and communications technology can facilitate.
Conclusions While many claim that the future will be dominated by biomedical technologies, these are, in the main, only tools to repair people with problems, whereas computer and communications technologies facilitate new opportunities, new products, new ways of living, and liberation of human beings potentials.
Sites to watch: Bob Cringely: http://www.pbs.org/cringely/http://www.pbs.org/cringely/ Peter Cochrane: http://www.silicon.com/comment/petercochrane/ Ian Pearson: http://www.futurizon.com/http://www.futurizon.com/ BT Technology Timeline: http://www.btplc.com/Innovation/News/timeline/statictimeline.pdf