2010 Trends: Alternate Computing Emergence and Convergence
The year started with the Mother of All Disruptions as the world teetered toward economic and financial collapse. The technology industry withered in general due to lack of demand. Intel, for example, reported its first loss in 21 years in the second quarter. As we head in to 2010, things seem to be on the mend, albeit slowly.
I thought I’d jump on the new near “top trends” bandwagon and provide some observations of my own for information technology for development (ICT4D).
Netbook fever and 1:1 computing in education begin to fade into the background
Ever since Nicholas Negroponte launched the One Laptop per Child project and Intel followed with the Classmate PC, the buzz has been about netbooks for classrooms, or 1:1 computing (one computer for each student).
The reality is that the majority of netbooks sold are not sold to schools, but to middle class consumers who are looking for a smaller notebook form-factor. In my 2009 travels, ministries of education in Latin America seemed to be the most notebook centric. Peru had purchased 150,000 XO laptops. Chile wouldn’t even consider anything that wasn’t mobile. As governments’ emerge from budget lockdown, I predict that they will look for more affordable and realistic options, such as PC labs and desktop computing.
Alternative computing models “cross the chasm.”
A desktop PC or notebook computer has typically been the primary way people in the developing world get exposed to computers and the Internet. That is changing rapidly with the introduction of solutions that significantly lower acquisition and maintenance costs and provide increased energy efficiency over a standard PC or notebook. For example, the company I currently work for, NComputing, sells a product that allows up to 30 users to share one, inexpensive desktop PC by hooking up additional monitors, keyboards, and mice to small access devices and costs about 75% less than a PC and uses 90% less energy. In 2009, NComputing reached 15% of the US market desktop computers in K-12 education.
Microsoft has also embraced “shared computing” for education, announcing a new product called Windows Multipoint Server that will be available later this year. Many developing countries, such as India, Brazil, Pakistan and others, now allow these type of solutions to be bid in addition to standard PCs and notebooks. Just as shared access will prevail over 1:1 computing, virtual desktops will become an increasingly popular option given the tremendous cost savings over traditional desktops.
Mobile phones and Computers
The final trend to watch is whether one form factor – the mobile phone or the computer – will win out over the other in ICT4D. With smart phones providing most of the capabilities of a computer, some argue this will be the ICT device that prevails. But is it really a zero-sum game? My opinion is that the computer and the mobile phone will coexist for the foreseeable future.
Sometimes you just need a full-size keyboard and monitor for an application. And sometimes you just have to be truly mobile (and by mobile I mean being able to transact on the move vs. sitting somewhere with a laptop). At Intel we often talked about “three screens” … the small screen (handheld), the bigger screen (computer), and the biggest screen (TV).
But all of these trends should lead to increased development through access to innovative ICT solutions and services that could be created and driven by social enterprises. I’d love to see a special report from BusinessWeek and The Economist on the convergence of these trends and its impact, but if not, we can always blog about it.
Thin Clients are a good innovation, however the high prices of Flat Screen Monitors, have made them expensive, hindering its penetration
Cavin, They can be remarkably cheap nowadays, especially for large government tenders. On tenders my company has been involved with we've seen 15" LCD monitor prices as low as $65. You can get 17" for $89 and 19" for single-unit purchases off the internet. They can be cheaper when bundled with a PC as well.
Thanks for this information, however you have not provided any information (url or otherwise) on where these $89 or $65 LCD monitors can be purchased, i hope you are not talking about used/reconditioned LCDs which have no warranty, and other associated problems, everyone is aware of.
Cavin, the $65 price I mentioned was for new monitors but for a very large deployment (>60K seats) so you will not find that price listed publically. The $89 price I found on TigerDirect.com … new 17″ monitor with a rebate. Here is the link: http://www.tigerdirect.com/applications/SearchTools/item-details.asp?EdpNo=4008589&Sku=H94-1706
Cavin, the $65 price was for a large tender (>$60K seats) so that obviously isn't a public price. The monitors were new. The $89 price was on TigerDirect.com after a $70 rebate for a new $17" widescreen monitor. Street pricing in China where I am now is around $125 for a 19" monitor. Go ahead and shop around and you should be able to find similar pricing. Bottom line is that monitor prices have come down significantly.
$125 for a 19'' monitor is for what brand, and how long will it take before it dies, now if you have the monitor price higher than the OLPC, is there a choice to make for a decision maker. Thin Clients are good for banking institutions, College Campuses, where data and software are stored centrally, but for developing countries, it will take several years to see them in a rural school.
We use thin clients in rural schools in Mozambique, quite regularly. We have deployed them (mostly WYSE models) in rural schools and community ed centers. We also have OLPC pilots in primary schools. Essentially, the infrastructure, conditions and characteristics are comparable in the two deployments here (OLPC and thin client labs).
Although it's true that what you refer to as the "central failure point" does present problems, my experience has been that the OLPC pilots *always* present more failures, hiccups and overall technical delays. With the high attrition rates we see and ever-fluctuating (though generally increasing) class sizes, the 1:1 deployment in general is much harder to maintain as truly 1:1 and as being consistently effective. At least with the thin client labs, we have a far better idea each day as to which machines are working and which are not, and lesson plans and lab times can be carried out accordingly.
I don't see the relevance of whether or not the monitors have a warranty: there is no comparable warranty or effective tech support for the XO where we are.
Lastly, just from a standpoint of efficiency: the students in these schools (and in other countries I've seen with OLPC deployments) waste a lot of time with XO problems. The touchpads become unresponsive very quickly, and in the high heat the battery life drops considerably after a few months. Where electricity is scarce, it is always best to get the most out of the little power time we can manage. The thin clients are simply quicker and more stable, given that we can arrange a small 12v battery arsenal with a few solar panels and inverters to keep the terminals going for most of the day. Children don't lose time running around trying to find an outlet to recharge, and they don't get distracted by the frustrations of getting their screen's cursor to do what they want it to.
"…my experience has been that the OLPC pilots *always* present more failures, hiccups and overall technical delays."
Were the XOs and thin clients used in the same way?
I expect that thin client use would be limited to a few hours a day, max, and only inside school. The strength of a 1:1 deployment is in the children using the laptops at home.
If the laptops are only used for a few hours a day in class, there is little rationale for a 1:1 deployment over thin clients in a IT lab.
They are of course used in many different ways, and context is only one of the reasons why.
The home use of the XO is one characteristic of it, yes, but even with home use, over time we are not seeing much "critical development" of the use of the software. At home they repeat the games (memorize, draw), and take more pictures.
Battery life at the thin client lab is 6 – 8 hours / day, the panels charge the batteries, which then go on to power through the inverter well after dusk.
If the delays in using the XO were not so frequent and time-consuming (flimsy touchpad, system crashes, etc), I would be more inclined to invest the project in them.
very good to hear about the project in Mozambique, i wish you could give us more concrete information, how many have been deployed, etc. this would be useful. Also inform us of the cost. The issue of warranty is very important, because it means the manufacturer has to adhere to high standards well aware of the dire consequences of producing an inferior product. I think you are aware that most products with no warranty have a short life span. Its also interesting that you are talking about warranty and technical support, these are apples and mangoes.
I am not marketing OLPC but your observations on these laptops are surprising. "In the high heat, the battery life drops considerably". Now this can be addressed by using them optimally, not on the beach. In conclusion Thin clients are good and OLPC is a good technology, an argument that one is better than the other holds no water.
The OLPC has significant advantages over thin clients, but thin clients also edge out the OLPC in some areas (despite these being very limited) . If i was tasked to make a choice for a low resource school, i would definitely opt for OLPC. If i am making a choice for an Urban School project with kids who have Desktops/laptops at home i would choose thin clients.
Keep in mind that what I'm talking about from NComputing technically isn't a thin client. We call them virtual desktops. Our warranty is 3 years, but failure is rare given no moving / hot parts. They are closed, durable boxes the size of a cigarette pack. The only advantage for OLPC is really portability, and with virtual desktops at 75% of the lifetime cost of an XO, it questions whether it is better to use those $'s to put more kids in front of computers then to give less kids portability.
Portability is a huge advantage not only in achieving 1:1 for students — the value of which is itself open to question in developing-country schools — in terms of enabling flexible configurations of _in-school_ computers. But the question you raise and the question of 1:1 in specific environments are both testable and should be the subject of research. One of the major barriers to effective tech use in education in developing countries is that education-research activities are minimal. Without independent investigation of these questions, and others, how could anyone make a decision instead of a wager?
I should amend that first sentence: "Portability is _potentially_ a huge advantage." Again, it's a testable idea but without the test.
Yes, research demonstrating this value vs. additional access for more students would be useful. I also think it should be measured on a scale of one's age/grade. I would guess the older / higher grade a student is, the more valuable portability becomes.
"and with virtual desktops at 75% of the lifetime cost of an XO,", please explain how you came up with these figures
Total cost of ownership (TCO) calculations. You can find a discussion on an article I posted to this site here: /individal-and-communal-c…
I reference 60%, but we've seen savings on actual deals up to 75%.
I have read the post, however the excel file is missing, your calculations are based on simulations and not live data, so its just a hypothesis. You can claim 4%, 89% or any figure, but you have no data to back it up.
We typically deploy 4 termanals to each server, which itself is a laptop. For a lab of this set-up, including solar panels and battery arsenal and cable, roughly $3500. The WYSE warranties are for the 3 years, laptop warranty varying by design. We typically use Dells than can handle running Windows Server.
But I reiterate that warranty and tech support are not "apples and mangoes" as you say. The warranty is an indicative comfort, but realistically no manufacturer would be replacing or validating a warranty on a machine put in this environment. The actual benefit of a warranty is much like tech support: it exists on paper only.
It is not comparing the technology itself that helps students at the end of the school day, it is how they made use of the technology. In rural areas, where teacher training is low and access to support nonexistent, the thin clients provide a better service, simply put. I would be interested to learn what you mean by "using them optimally" in regards to the XO, because we have tried everything. What do mean by "optimally" exactly? That which engages the children and excites them about learning? Because their favorite programs (Record, TamTam) are those which use up the most battery life. Should we have them doing rote copying from the chalkboard on to the XO, contrary to student-centered teaching, just so that we can "optimize" the battery life? Seems antithetical to the project's purpose.
I should note I wish we were on the beach, but no…just in the middle of a serious drought!
1:1 computing might not have realized it's potential or anywhere close to it – but this might not be because of the potential of the platform but rather the difficulty in ensuring that on the ground we really (with a difficult infrastructure situation) to realize the full value of the tool to improve learning outcomes.
1:1 computing gives us a tape player let's say. We need digital libraries on the school server, realistic techer guidance materials, teacher training and pedagogical models compatible with the local culture, a whole ecosystem to really make it all accomplish what we're looking for.
The issue that I see with lab or school machine based models is that they do not so readily help overcome the biggest obstacle to learning objectives – lack of education time / teacher time. Parents often can't provide guidance. Schools are taught in shifts. For what should be a 1 hr lesson you might have only 25 mins. Having 1:1 computing you can effectively extend this time using blended learning methods / planning. How do you do that thin clients etc?
1:1 computing also provides us with the potential to provide huge reams of content even without content by just connecting a decent size hard disk to the school server, that kids can then study in their own time according to interest, how does that happen with lab models?
Lab models I'm sure can be effective for introducing computer literacy; but given that time will likely be so limited there they will probably teach mostly computer literacy.
Also huge amounts of computer in the school all in one place overnight would be quite possibly more of a security risk than laptops dispersed at various homes… Given that the guy watching the school is unarmed / poorly paid and it takes maybe $15,000 to get yourself smuggled to Europe… well one can see the problem there.
I personally think thin clients are an innovation more appropriate to those applications that lend themselves to shared usage and where secure overnight storage is not a problem, particularly things like local government / small business / NGOs.
Mike, I think 1:1 computing is clearly better than shared access computing. But realistically most governments don't have the budgets to put computers in the hands of every student. And if they do try to do 1:1 computing, it just reduces the number of students that have access. And the "thin clients" that I am talking about here are not truly thin clients … we call them virtual desktops. They allow students to share a single PC (with their own monitor, mouse and keyboard, using the computers excess performance capacity. And ironically, the only country that has truly done 1:1 computing is Macedonia, which they have used NComputing's virtual desktops to do it within a small budget.
please explain the difference between thin clients and virtual desktops. There is an interesting argument, that OLPC will not achieve 1:1 computing, which is not the aim for OLPC. Even if you try to achieve 1:1, are you saying its easier to achieve 1:1 computing with more expensive equipment (read Ncomputing/Thin clients/Virtual Desktops), and harder with cheaper alternatives (OLPC).
Thin Client definition: http://en.wikipedia.org/wiki/Thin_client
"Ultra-thin" client or "zero client" mentioned in this link is closest to what I consider a virtual desktop. Thin client is different than a PC in that the CPU is lower-end, has no hard drives or optical drives (uses flash). Costs are the about the same with a PC. Virtual desktops have no OS, no CPU, no HDD, etc. They just provide an interface for keyboards and mice. Software is loaded on a PC that allows multiple virtual user sessions to be distributed to multiple separate client devices.
You can get many of these questions answered through Google or Wikipedia.
Its interesting that in talking about the difference between Thin Clients and Virtual Desktops, you talk about the difference between the thin client and the PC. In your earlier post you mentioned that NComputing's devices are not thin clients, and they are actually virtual desktops. And that was my interest. There was little need to go to Google or Wikipedia, to explain your choice of words, which you have failed to explain.
So we can disagree, but i will call them thin clients, because that's what they are. You have consistently avoided to mention the price of these 'super devices'.
I will conclude by saying that the principle behind this super technology (Thin clients) is very good but the high costs involved in the acquisition will mean, they will be relegated to isolated installations around the globe (read Developing countries).
Hi,
While netbook fever has certainly passed in commercial markets, it's difficult to tell in education if it's passed or yet to hit. Netbooks have been on market for too little time to have penetrated the project-planning and funding cycles of national governments: XO-1 was announced in 2005; eeePC was introduced in late 2007; developing-country commercial markets received netbooks somewhat later than those of OECD countries. And OLPC is the only netbook OEM, more or less, to market to ministries of education.
Netbooks present an array of features that could be really important in developing-country education systems: most netbooks don't have hard drives, so they're possibly more durable and power consumption is low. And they're cheap. The importance of these features is magnified in large-scale projects, in education systems in which personnel lack ICT training or familiarity, and in infrastructure-poor environments. Netbooks (as do notebooks) support many different configurations–several netbooks in a few different classrooms, one netbook in every classroom (and a projector, perhaps? An LED projector?), or a bunch of netbooks in one classroom.
This flexibility means that netbooks (and notebooks, with higher cost, higher power consumption, and reduced durability) can support the teacher-led pedagogies that are what most teachers use: "I have the laptop, I have the projector, I show you stuff. At least it's stuff that's more interesting than the textbook stuff we had before. Now, if I also had a digital whiteboard…"
One netbook or a few netbooks in a classroom can support station-based learning or collaborative learning. A bunch of netbooks in a classroom can support computer-lab-style learning, say with students using educational software or productivity software independently.
Netbooks can enable all of these activities AND they can support computer-lab-based ICT instruction. While it might be preferable to eliminate ICT classes and instead contextualize the development of ICT skills in other subjects, a lot of developing-country school systems (all of the OECS countries in the Caribbean, for example) already have ICT curricula and exams plus teachers to administer these. And in a lot of cases (but not all) they've invested in technology. It's unrealistic to think that "basic ICT" instruction will go away without a kind of glacial resistance. At the same time, it's a waste of money to launch province-wide or nationwide programs to invest in computers in schools if those computers are going to be dedicated to learning technology-as-subject.
Netbook fever might not be out of the picture just yet. It's probably too soon too tell.
Hi Ed,
I agree. The wave is yet to hit. I work with 22 countries in the Pacific Islands and your comment about the planning and budget cycle is quite right. We started 5 small,scale projects to kick it off, but now have 15 countries signed up to projects. Another 4 or 5 will start larger scale projects this year and the balance the following year.
The cycle is much slower in developing countries, but that doesn't mean the interest or commitment is not there. Quite frankly, they all know they have to introduce ICTs into their education system, but they lack capacity to do this. OLPC is a neat package which makes the project much easier.
Ed, my point about the netbook fever slowing in the education sector is that I think the OLPC and Intel Classmate PC hype mission will have become but a memory (this is already happening). I have no issue with netbooks or notebooks for that matter for use in schools, but I question the tradeoff of the benefit of portability over cost. Higher costs mean less computers being purchases, reducing the number of students that have access. Is it better for more kids to have access to computers, or fewer that can now carry it around with them?
Its good you indicated that you work with Ncomputing, short of marketing their products can you give us the advantages they have over the OLPC. And when you mention the cost, please don't indicate the cost of the thin client, without considering the monitor and other gadgets, this could prove that Thin Clients are a good technology. I hope you also add the disadvantages of Thin Clients. I will only highlight one, which is having a central failure point.
ED Gable..
I am in total agreement with you on the usefulness and relevance of Netbooks. In fact we are promoting such concepts for developing countries where there areas there is no electricity. We are recommending our solution to close the digital divides and strongly advocate this use of Netbooks with low cost solar power because power consumption by Netbooks are low.
The overall cost to implement is much cheaper than traditional ways using desktops or notebooks.
Our solution advocates the use of usb drives too as one single pen drive can accommodate years of tutorials and execises.
Being cheap , new and low cost of operation, is perhaps the best solution for developing countries to role out for the entire country.
Currently we are piloting our solution with our Government's telecenters and now providing our tools to empower teachers and contents like English, Maths and Chinese Mandarin free for all schools around the country and globally.
We are even using our tools to help out poor Haiti the moment we knew that UNEXCO had announced that
“Education is at the core of Haiti’s recovery and is the key to Haiti’s development”
http://www.unesco.org/en/education/dynamic-conten…
We are trying to provide the entire Haiti free contents that all their schools can use to jumpstart education at a fast pace because our solution is able to reach out to the most remote without expensive broadband.
We do advocate the use of Netbooks and low cost solar power because by now Practically the entire Haiti may be without power not only in rural areas but the cities as well.
They are the most green too using that pair .. Netbook with solar power. At night they act like a lamp.
We have just started our 1 fo 1 or 0 for 2 program .. where we get donors to buy one and we give one to Haiti or donors pays for one and we give two to Haiti. This is on top of our free offer to all schools and universities there a free set of our tools and entire year or more of contents free.
If anyone of you wish to check us out, feel free to visit us at http://www.paperlesshomework.com
Yes, I do certainly advocate the idea of a netbook plus an LCD projector(if there is power supply) for the entire class. That at least can fit into the way our classes are conducted. 1 for each child? No the design and layout of the classes do not fit in. Too messy. Teachers/students cannot even concentrate.
This anyone with a full set of computer lab will know… dont have to conduct a 1 pc for 1 child to know.
Regards
Alan
The AGE tools are good but its better if they are customized for a particular country, different countries study different things at different times, and while the content might be similar, it will not be useful, if its not customized to the unique situation of each country. With no chance of ever having a global curriculum unlike for higher levels of education like Universities, customization is key.
Cavin Mugarura,
I do agree with your about contents being customized for a particular country. That statement has been forwarded to us a "thousand" times.
What we are doing is very unlike most content providers most of us are familiar with.
Most providers will provide the contents or the tools or the contents management systems like Moodle.
We provide all three.
1. The content we call the base contents
2. The tools to create contents by teachers (so contents relevance is no issue) or by MOE of the country concerned
3. The complete ecosystem of tutorials and homework assignments from the point of teacher creating it to the point of collating the students' scores .. all completely without paper and without the need for Internet or broadband.
We provide all 3 in 1.
So when we sell to a country… we tell the government. We have the contents you can first use if free. You need local contents we provide the tool for you to create it. We provide you with a simple easy to use contents management system even a standard one child knows how to use.
On top of that … it costs almost nothing to achieve many of the tough issues like collating results nationwide without spending a bomb.
With our solution therefore the issue of localised contents do not arise at all. It is expected that the thousands of teachers having been given the free tools to create contents would easily localise the software.
This is why we say,… it is ideal for unfortunate countries like Haiti or poor countries without the budget for broadband…but do have a bit for netbooks. Yet once nationally implemented, the countries' ICT usage index for education by students can equal developed countries.
It is hoped that by the end of the day, there would be available thousands and thousands of contents freely downloadable anywhere in the world .. even in areas with only slow connections speeds. We share the contents created.
In fact anyone wishing to have our entire contents to be placed in their web sites to create more traffic, we can give permission too. More traffic means more revenue to sustain.
That is why we are giving our tools and contents free to all schools, kindergartens, universities and telecenters. Check out our Mandarin modules… with voice and run offline and small small you would never believe. A video is there to prove it at our web site at http://www.paperlesshomework.com. Compare using our contents to using books to learn a language. Most such contents must be online and should have broadband. Not all have such a luxury.
Alan