 This is a picture of a school in rural, middle of nowhere Cambodia, it's about six hours The cost of the laptops are really expensive and the cost of the fuel for generators are really expensive It's one thing I'll start with, I want to try to reset people's view of the world here The first problem was when the kids were giving laptops was persuading the parents to let them use the laptops when they were taking them home Because laptops were worth more than the houses that they were going home to Once they got past that point, the next problem, the next realization of a day or two later was For most of these kids, this was the first time there was any artificial light in their house That's the bulk of the world Most kids go home at night to houses with no electricity and no light Okay, so this is sort of what I'm going to cover in this talk In some sense, this being an E-audience, I'm going to try to adjust the time to you why we designed the machine the way we did We've got the luxury which learnings people typically have, and free software people have not had Of actually being able to design around hardware from close to scratch This is a lot of the set of freedoms of design that you typically don't have And it's a lot of the still directions that you normally can't do in the Microsoft Ecology At least not in most your Microsoft There's realities about hardware design that a lot of people don't fully appreciate It is very different than software There are realities of what ingredients you can actually get to make your sausage And you can only make as much sausage as you can actually get the ingredients for It isn't like software where you can always run off a few million more copies It doesn't work that way Some parts of the other recipe in the substitute that said those are called second sources But other parts you can't, they're simple source players And if they stop being able to make your component, you're in real trouble If you make a lot of sausage, you can actually get some custom ingredients made And depending upon how long in advance you start, depends upon how unique and new the ingredient can be So full custom OSI is like a three year early time You can build an ASIC in as low as three or four months if you know what you're doing You can actually build a couple of applications to the ICs and the machines that we have here This is the so-called B-Test 2 machine, it's the second beta test build we've done So in any case, what we're out to do is to educate kids Or help them educate themselves, that's what we're about We're not about free software per se However, you will see how the two go hand in hand in this talk There's something of order one to one and a half billion children in the development world This is a lot of kids So if you knew the math, how many machines per year should we be doing If we could build a machine that lasts five years How many laptops were built last year? Does anybody know in this audience? I think it's like 65 million, 70 million, something like that You can see that if we do this for real, we should end up building a lot more laptops than everybody else put together This is the reality of what a school in the capital city of Nigeria looks like This school is fortunate, it has electricity in one room There's no lights or nothing in the regular classroom And that's the capital city, you go outside of the capital city of Nigeria It doesn't get better, it gets worse Another picture of kids from that school I believe Another thing I want you to think about is again, I had breakfast with someone this morning Who had never focused on the fact that most kids only get, say, five or six years of elementary education So the people we are trying to serve, the kids we're trying to serve are the young kids Not the middle schoolers, not the high schoolers, to which we have been primarily catering We're trying to fundamentally help kids who are learning to read and getting their help They're very basic education And in some parts of the world, however, what we consider basic education is quite interesting I was talking to by the name of Greg Gortenson, who's been learning Swills in Northern Pakistan And by the time those kids leave fifth grade, they have learned five languages Quite impressive But yes, in any case, if we just continue the way we're going That's a risk-quite assault going up in the risk Computers and education in the developing world is quite a role That's in Apple too, Seymour there, who unfortunately was badly injured a few months ago This is, I think, from, well, whenever Apple took its recurrence, that would be late 1960s or something like that So this is not a flash in the pan The difference is we're trying to do this at scale because we believe that there are network effects That take place when every child has a machine We're called one-by-one proper child for a very good reason We don't believe that sharing machines or computer labs are an effective way for kids to use them Do we share our pencils? Very hard to use it as a basic, as your book If you're trying to share it, it's hard for two people to read the same book at the same time So we're really mean when we say it When we say one-by-one proper child, it's a basic premise of the project Okay, so there's lots of interest from all over the world We're likely to launch in the countries that agree here about Thailand's nifty courtesy of the change of government That always makes things much more unrepeated Governments are not necessarily safe Our premise is that kids all over the world, and this is true so long as they get a decent nutrition Are just as bright all over the world as they are here So we think that if we can get machines into the hands of every kid That this is a really fundamental tool This presents a huge set of challenges, of course you might imagine Another point of view is we don't believe that kids are just passive consumers of quote, content From the content providers of the world We believe that kids are creative Those of you who have children, almost certainly your children keep journals At least in their early years As you may know, reading and writing are intimately related And so writing a journal is something that's fundamental I'd watch my own daughter now in 56th grade be able to actually write English reasonably well Which is quite remarkable This is a program called TAMTAM, it's under development This particular version isn't out yet But the intent is for kids to be able to listen to our music The other thing is we want kids to be able to experiment This funny gizmo on the left We did this for a demo of the so-called A-Test Awards at a conference last May That's a photo-cell at the end of a wooden spoon The photo-cell costs less than a dollar This allows you to conduct music Because you can detect what the light is And figure out what the rhythm that you want to get Despite detecting what the light is So we want it to be very easy to put things in One of the unique things about this machine is that the audio port can also be used just for DCM You can use it as an ATV converter rather than just for audio The network is really fundamental We want the kids to be able to share, we want kids to be able to work together If you watch your own children with their friends And often see two kids in front of a machine Visually teaching each other about the particular thing they're playing with Similarly, we want the teachers to be able to work with the children So the network is really fundamental However, this again brings a whole set of huge challenges There is no infrastructure in many parts of the world How do we get the network out to the middle of nowhere? So we believe it's just as important that the kids communicate with each other as with the internet We want them to get out of the internet in a serious way So the intent is in fact to get internet connectivity to all the schools Depending upon where you are, there is either good or terrible connectivity Pylen, for example, has 30,000 schools They claim to have connected all 30,000 of them On the other hand, if you go to Nigeria, you get a very good different answer Our point of view is that you can actually build networks on your own very cheaply, if you know how And the other observation is for a lot of things like cell phones The bandwidth that isn't being used goes onto the floor So there's interesting things you can play by government twisting arms of cell phone carriers To allow the bandwidth that's not being used for regular paying phone calls To go toward internet connectivity This is something in the middle of nowhere, and I think it was Pakistan About some years ago, setting up a network This is going across the student border from Pakistan into Kashmir, if I remember correctly These kind of radios are now very cheap, you can build them out of it as well This is a picture taken of somebody hacking away in downtown Athens It turns out, making their own network out of cheaply opponents So what we're intending to do out of all of this is actually ship what's called mesh networking The individual computers are able to forward packets on behalf of nearby computers You also notice that we actually have antennas that stick out That makes a significant difference in the sensitivity of the radios James Cameron has been doing experiments in the Outback of Australia Which is not as good as it ever gets So it's at least one extreme of what you can do with ADO211 Even without directional antennas, when he's had line of sight I think it's nearly two kilometers, it's been really amazed by I don't think that would be the more realistic thing of course, it's sort of 400 meters And then of course, you start thinking about meshes and map meshes That get interconnected by various point-to-point sorts of lengths And then you have to backhaul however you can back to the internet So that's the kind of architecture that we're working on I'll talk a little bit more about how that's driven, the hardware design here I'm going to just go into what our design trade-offs have done Since this is not the typical laptop Obviously, we care about it being safe And you really don't want to have a 3kg laptop on a 6-year-old It needs to be pretty light We also have a feature that can be smaller than usual Because guess what, kids are a lot smaller than we are Kids like things that look like toys It has to be low-power Some people think, oh, we should ship all the laptops and board desktops We don't use any more here to the evolving world Well, as I demonstrated before, the numbers don't work just in terms of quantity And then you think about the power plants required to power these things Dimensional laptops are often consumed 25 watts You should be buying laptops that's 25 megawatts Is this wonderful? You really need to worry about things like power and scale Obviously, we have to make these things cheap If we're going to be very expensive If we're going to be able to build them in any sort of quantity So our big challenge is our infrastructure How do we get what kind of power, how do we get connectivity? Obviously, as Thailand has some correction been demonstrated Political instability is also a major challenge We have to worry about physical environment Kids don't get on school buses in the developing world They walk home We would like the machine to survive getting rained on At least when closed So we worry about that sort of thing You will notice it has a rubber keyboard It isn't that the keyboard is less expensive that way But that may be true We have to worry a lot about water and dust So the conventional keyboard will not survive In any case, there are other indirect costs we worry about If our software is big and bloated, we have a problem I think this has shown how big and bloated things can get We need to fight this tendency hard in months I believe people are beginning to take this more seriously So now let's get out to where the costs go in most laptops If you try to make a brilliant and extensive laptop It turns out that sales and marketing would be your biggest single expense So basically we have Nickel Snack or Ponty Traffin in the world As our president of marketing He's quite an amazing force of nature, shall we say Then there's a lot of expense that goes into supporting big, bloated software And another major cost is the display The display on your machines are expensive As you'll see, they also won't even work We need to be able to have a display work Kept a little hot pot today It often consumes 20 or more watts I think one or two watts But something a little over a year ago We got together to really go for our first major designer review And understood that building a machine which looked like A steady state power of something of 1 or 2 to 3 watts Was not good enough And the reason why is reality at the first point A small kid cannot generate much power We want to have a situation where The child can rely on power being available By being able to generate it themselves With some sort of a hand generator It turns out that a kind of new machine is a bad idea But there are several generators being made As external devices We really want the kids to spend most of their time Actually able to read and write and like So our goals have been more like 10 to 1 If the child has to generate power We want them to not be spending most of their time Generating power for the machine That sort of thing There's another observation we figured out Which was the connection at work demands of the wireless Beyond preferably all the time That's an interesting problem If the CPU is on all the time And the CPU is taking a couple of watts The numbers don't add up very well So, and in particular If the child is not confident Will be power when they need it The first thing they'll do The inventable features that they are Will be to try to disable the wireless So they'll have power when they need it So we finally realized that Most of the time when you look at your laptop Not much is going on The screen is being refreshed And we're forwarding packets And that's it Most of the time your CPU is doing nothing Why is it on? Tell me again Why is your processor on? Why are the processors in the audience on now? How many of you are actually using your machines this instant? I don't see any chance to go up The wireless Well, yes, the wireless isn't working She was actually in the audience So, the sensitivity of her mother was mentioned So from our point of view We want to keep the processor off as much as possible When an institution where there isn't much power available And leave the wireless able to work And possibly the flight audience Will be reading something So how do we solve this problem? Well, there are two solutions to this One of which One piece of it is what's called a decon check Which allows us to leave the display on The LCD on And turn off the processor board So we can have the screen on And turn off that CPU Which even on all of our CPU Is still consuming a few watts That costs us I don't know Something probably five or seven dollars Like that by the time all of a sudden done Was a small RAM chip that we had gone through this flight We also made a A limited decision to use a wireless chip That nobody had heard of Made by Marvell It actually has an R9 processor in it And about a quarter megabyte of RAM And so it gets able to forward packets Even if the CPU is off Okay? That's why we selected Marvell chip There is no other chip like that on the market right now So At the end of the day Depending upon what you're doing These people be down the Sort of one half to one and a half watts Area depending upon Whether you're using the screen Or wireless, that sort of thing So we think with pretty few watt hour batteries We should have pretty long life This is a typical number obviously Obviously A piece of this is how quickly Then we wake up the machine If we want the machine to appear to be alive And we have people who are suspending that right and left We're going to have to wake up the processor quickly It turns out that Geo is not ideal It takes about 25 milliseconds To do the wake up But it takes about a frame time To get this playback from the Decon chip So I think it will be about 15, 70 milliseconds sort of things So this is just at the edge Of human perceptibility I did some measurements on And I showed that From the time that it came out of Reset To when Linux was able to schedule Processes Took about 10 milliseconds On a 200 meter bit slower When I've asked the current community How long does it take to come out to suspend It's been It was a great fun trick question For a big part of last year But the typical answer was In the one to second range Of the reality is It should be measurable in a little seconds Okay So there are other things we'll do Like drive the channel slowly whenever possible This isn't particularly novelized Where people are doing this sort of thing Obviously it changes the rate Of which we update the screen Depending upon your doing it The screen itself is very novel As you'll see As I was talking about before We're using this from our bell chip Or you felt about it We're Our bell is implementing A standard new development Called 802.11.S So it's Able to do Forward packets Without the processing on it all We're just bringing this up right now In a more serious way That's what was winning last week We Showed it was technically possible To do by about eight or ten months ago Not only can it happen for real We obviously Care about power management In a really serious way Most notebooks are often taking Ten seconds to wake up I want to wake up under A hundred more seconds I think that's really cool So the Other thing we obviously Worry about was cost So this is really Like a consumer electronics Subjection mold of plastic You need volume To afford to make the Plastic Injection molds They're ridiculously expensive By far the cheapest way to make A Thing if you're building it in high volume But you can't afford to do it Unless you are making things in high volume So if you have a chicken and egg problem You must have large volume In order to get prices down So We're also Getting rid of most of the distribution costs Basically What's happening is the countries Will be ordering directly from Ponca Which is the ODM involved For those of you who don't know The end is most laptops In the world are actually built By a set of companies Who you've never heard of Ponca computer is the largest of them One out of every two laptops In this room Of almost any manufacturer Who you care to name is made by Ponca So we have the 1000 comic Roller working with us It's really quite fun to see How that operation works Oops I hit the wrong button by accident Okay, so we think the power Will be way down The 12 watt worst case Is if you have all three USD ports The more typical things that will be Down in the The lower end of the store lessons We're doing It's not a very fast processor It's sort of what we all have For five or seven years ago It's a GFGX2 We have 128 mega-gram Thankfully OpenOffice is not something You want to inflict on an 8-year-old Anyway So that's not a problem So in fact most software just works There's Megabyte of serial flash For the Firmware, I'll get to the moments The USD ports is an SD card slot The wireless Audio Is also a camera and a machine And half a megabyte Rather a man flash Using Dave Woodhouse Of GFGX2 files Data compression So for typical programs in data It feels more like Something about quite set size We also have a new Touchpad Whether we're actually going to ship this or not I don't know yet because we're still slightly over Budgeted We may pull it out, I don't know But in fact The one here Which I have either being used as a Dimensional touchpad Or very easily used as a stylus For learning how to write Strangely There's a VGA camera It's the kids of built-in pictures It's the green So I'm glad we have scale luxury Bright Okay, so I want to know a little bit about power systems design This machine Will take more or less anything Nominally between 12V and 24V You can plug it directly into a Power battery. This is like where you need One of these Ricks That are expensive That produces exactly 18.5V Because guess what Lithium ion Really likes exactly the right voltage To charge And we know what happens to lithium ion If something goes wrong these days As Ellen Cox demonstrated So we are not using lithium ion Our Our original intent was to use Nickel metal hydride We're also going to be using another chemistry That's just becoming available Which is not lithium ion But another lithium Battery chemistry That is similarly safe And ecologically friendly Lithium ion also a lot to be Re-circulated We're trying to make We worry a lot about what the lifetime Gives The notebook manufacturer sees batteries As a profit center And so they're perfectly happy to have New battery every year Because we can charge them 300 or 500 Comments, aren't they That's not adequate For us So we worry a lot about Making sure that there is cost Of the charge with battery many many times There's a gany charger That we just got the first samples With of Has a power battery And you can charge 10 batteries at once That sort of thing There's also solar cell panels That will come in But also the ultimate Fallback for the kids Is generators There's a crank being built It's also a pull cord device Looks very promising If you can make it What a robust That It's a former media lab folks Who started a company called Botanica To build That's much more efficient than a crank Your crank only uses the little Muscles in your arm Pull cord You get to use your shoulders Or you can use your leg That's a much better idea The display itself It is a New and novel display Costs maybe 40% of what a animal Of this size would normally cost Most of the cost of an LCD Is not the LCD Most of the cost Of what you have in front of you Is in fact In the three color filters In a fluorescent Light bulb which is high voltage And therefore requires an inverter And that costs money And how to drive the electronics What's more A conventional display Does not work in bright sunlight And most Many are most kids in the developing world Got caught summer all the time Out of doors So we have to have a display That could actually be used In bright sunlight So conventional flat panels Even if we could afford them Which we can't Don't work Meredith Jepsen With a new form of LCD In color mode It's backlit And we have this funny pattern you see Red, green, blue, red, green, blue And so on In Front look Ray scale mode It's a 200 DPI display So this thing Is 1200 by 900 resolution So we have a lot more pixels on the screen Most of the laptops in this room Have people to be able to read books nicely That sort of thing In color mode Depending upon the ambient lighting The resolution is anywhere from about 800 by 600 If there's no ambient light Remember in this situation You actually still have Intensity information occurring at every pixel So the effect of resolution Spatial resolution Is not a simple There isn't a simple function I just meant I went through the pixels Therefore it works out to the following resolution So as the ambient light goes up The color saturation goes down But the effect of resolution With this light is going up This is a complex situation However, if you do test patterns In fully dark room with backlight The worst case is around 800 by 600 Sort of resolution In a lot of other environments It's effectively somewhat higher The power consumption is very low We do not use The fluorescent light bulbs We use LEDs They are color matched LEDs But we don't care that the color fidelity Is exact And so we can take basically All of the manufacturer All of what the LED manufacturer does So cost us a dollar or two For the circuit 12 LEDs Which are also replaceable Because part of the problem is In places like Libya The lifetime of the LEDs At elevated temperature It's expected to be something like Three to five years So they need to be replaced once During the life of the machine And that's possible to do So we can actually replace that Backlight if it should have failed So at worst case This display consumes about one Lot of power Most of your displays in the audience Here are taken between In grayscale mode At 25 hertz refresh This display takes about 100 mole watts So this is a fundamentally Different flavor display Than we used to As I said Part of the trick of making Our power budget is being able To turn off the CPU Obviously we have to leave the power For the DRAM being refreshed But not active So there's usually a chip Between the CPU and a TFT Called T-Con We have an megabyte of memory And double recalls of the D-Con And it drives the funny pixel Thing for us It also can do some coloring By always saying That sort of thing So this is part of How we get to the point where We can leave the display on And turn the CPU off So this is sort of The difference between it with normal mode Normally be The other The other ASIC we built Is a thing called CAFE This is called Slighting down the slippery slope The The Geo turns out to have A man flash controller which Always say sucks So we have to do something And once we have to build the chip To interface to NAND Ourselves We also The country's wanted A standard of flash So we have a SD And also a camera The dual mode display Is actually usable in sunlight And that for example is a picture Of a real classroom Out in the sunlight This shows Conventional IBM ThinkPad In sunlight and how much you get to see And there's our display next to it Okay So I'll quickly run through We did a lot of I guess we've lost some of our pictures My regular laptop's in the shop And I'm having real trouble This is what a production line Looks like It's the first machine Built in November I had fun watching the build The second build of Three or four weeks ago Of machines We went through a whole pile of pictures We're going to be able to see Different mechanical designs Slight So A lot of the mechanical improvements Will be in the next So-called beef test We're trying to make something That doesn't have sharp corners Or edges In case somebody falls onto it We're trying to make it as moisture And dust and dirt resistant as we can That sort of thing You will live in a slide One row And you have a shoulder One of two poles What the ultimate environmental difference Between using conventional textbooks With all the paper That's not great From the 20 day versus building Electronics is a very interesting question We're having some people look into this What the environmental differences are Making silicone is also interesting So I don't know what the wash is But environmentally it's not It's not obvious Okay, I'm going to move on quickly The software This is We care about open source We want the kids to be able to learn computing And the teachers to be able to do so We want them to be able to To bring those who are interested To relate good things We're obviously worrying about Software bloat in a serious way We're worrying a lot About system security Because we're building a very large ecosystem We hope to build five of them From the machines next year And that would be And make us a very big target The other problem is theft Of that in a way Son opened up What's called open firmware Without even press release So we're using that rather than Conventional BIOS We had news Linus's boot loader for a while But open firmware Ended up being a better bet Linus is big enough to Get everything shoehorned into one native Or one native like Flash And we care a lot About having complete control We're using Linus BIOS And open firmware for this This allows us complete control Over the resume path And so we aren't running through ACPI stuff Which would be measured with Tens of milliseconds Even if the BIOS doesn't do Other stupid things I've noticed BIOSes that seem to Want to sit there for 10 seconds In the worst case that some device Out of some USB might We also want kids to be able to see How the shoes really work There's a basic set of applications being worked on A user environment Remember that our user interfaces Have been designed for adult Developed country office workers Not 6 or 7 or 8 year olds Who are just learning to read Do you think this is a good match folks? Do you think our conventional Talks about whether they're No more kidding me or this stuff Or Macintosh or a good match for young kids Don't think so What we care about is An enabling collaboration Between the kids and their Teachers in a really serious way And a UI that's This approach that's comparable By the kids There's a concept of a zoom interface Where this is the local machine And you can see your buddies Everybody who's on the local mesh So we are bringing The facilities for collaboration And making collaborative applications Really very far forward In the UI So again this is The Different sort of modes This is a mock up of the journals Being built right now So here you In the upper right hand one You're seeing a bunch of Kids working around with common activities So whether they're browsing the web Or PPP or stuff like that If your application is big Slow and power hungry I'm afraid that you're out of luck We often have a lot of choices For any given tool We obviously care about small and simple So Abbey Word is very nice But it's too much for a 7 year old It's great for my 12 year old by the way She uses Abbey Word to preference To either Microsoft Office For a 12 year old That seems to be Abbey Word But it's just about the right thing But for a 7 year old it's too much So the Abbey Word folks are building A nice simple version for the 100 bits It was great And again there's also a collaborative aspect To Abbey Word who stays There's lots of ways in which we can Use help I have machines to give to people Who are helping us I even have a couple I'll probably give away while I'm here Some people question Why do we use GTK and Kendo Turns out that's the best localization ability We care about languages In a way that Language coverage In a way that is the way you know And That's really too bad Two things I really wanted to show Most I can't show I have this wonderful logo For people who Work on blood It has an overweight tux Very sad With the international symbol So this is Bloodbusters We follow ourselves Let's fight blood We want to keep tux Slightly touchy self While it was In the files it was In the presentation so It lost They have their problems They have their problems We care a lot about NVM services We think of servers As performance optimizations Rather than being necessary So for example our chat system We're worrying about uses of Or a protocol but Will work even when you don't Have a server accessible Because we want kids who are in a village Someplace away from the school To be able to still chat So think about that sort of thing When you're doing things We really can't have system management In any formal way Courtesy of things like a body In any cast and so on A lot of the manual So we hope to Or where we are now Is we just built our second Build of machines There'll be a third build later in the spring At which point we'll do Much larger education Trials And it looks like End of July That's when we should end up In high volume mass production And I want to End by saying thank you For this project Without your efforts Over many years Would not be possible So thank you all We're all nerds here Sixths for pixel The pixels are packed very carefully If you do If you do the math You can just fit that into a megabyte Good question A very, very good question But yes LCD is giving you In reality about six bits of ratio So that's the answer To how that works Barely Yes sir A lot of applications told me The question is How do you do localization If you're using mostly images In the case that Icons are not completely Transferable culturally They're more so than Text is But yeah I can imagine We're going to have to build different icons For it Understand we want the kids to learn To read and write and really do The machine has to be usable By a child who hasn't Who's just in the process of learning Right And building their Red shirt Yeah But actually right now they break off easily This is not a feature In the next build we're making the marbles About bend About bend enough to push them Around on the axis Like that again That's a very good question These were not strong enough For people who've been taking these machines And dropping them on the marble floors On the Yes on purpose We have several hundred machines After they went through Thermal testing and so on and so forth Ended up being tested and Told destroyed So we could figure out what needed to be strengthened Now a few of the Strengthenings are in this version The next version of the So-called V-test stream machines The results of that Work Making a more robust machine It takes a long time to modify Injection molding Molds For lean time on those Are quite long, measured in months So you can turn new You see for it it's much faster You can do a sheet metal much faster But new plastic It's expensive and it takes a long time Thank you all