Category Archives: Technology

New Server Uses 75% Less Power and Space

SeaMicro drops an atom bomb on the server industry

[SeaMicro] has created a server with 512 Intel Atom chips that gets supercomputer performance but uses 75 percent less power and space than current servers.
…
Today’s servers are so inefficient when it comes to being properly utilized,” Feldman said. “This misalignment between the server and the work load is the root of the power consumption problem.”

So SeaMicro guessed that servers could benefit instead by using lots of smaller processors, and it got lucky when Intel started promoting its low-power, low-cost Atom chip for netbooks. That lowered power consumption, since Atom processors deliver three times the performance per watt versus Intel’s server chips.
…
But SeaMicro also attacked the power consumption in the rest of the system, which accounts for about two thirds of the power consumed by a server.
…
it applied the concept of virtualization to the inside of a server. Feldman designed custom chips that could take the tasks that were handled by everything beyond the Intel microprocessor and its chip set. The custom chips virtualize all of those other components so that it finds the resource when it’s needed. It essentially tricks the microprocessor into thinking that the rest of the system is there when it needs it.

SeaMicro virtualized a lot of functions that took up a lot of space inside each server in a rack. It also did the same with functions such as storage, networking, server management and load balancing. Full told, SeaMicro eliminates 90 percent of the components from a system board. SeaMicro calls this CPU/IO virtualization. With it, SeaMicro shrinks the size of the system board from a pizza box to the size of a credit card.
…
This advance is coming just in time. Google said recently that if current power trends continue, the cost of energy consumed by a server during its three-year life span could surpass the initial purchase cost for the hardware. The Environmental Protection Agency reports that volume servers consume more than 1 percent of the total electricity in the US—representing billions of dollars in wasted operating expense each year.

Sub $100 Tablet in 2011

I must admit I am skeptical. If it happens this looks very cool.

One Laptop Per Child Revamps Tablet Plans

On Thursday the foundation announced a partnership with chip maker Marvell to collaborate on a sleek and cheap touch-screen tablet for developing-world school children, a device it now plans to launch at the Consumer Electronics Show in January 2011 for less than $100. One Laptop Per Child (OLPC) says that’s close to two years ahead of its scheduled release for the so-called XO-3, the long-awaited upgrade to the non-profit’s XO, the so-called “hundred-dollar laptop” launched in 2007.
…
The first XO, for instance, never reached its price target of $100; it now sells for $172. About 2 million of the devices have been sold–a significant achievement for a small nonprofit, but far less than its initial projections. And a flashy double touch-screen model known as the XO-2 was quietly scrapped last year when OLPC decided it couldn’t be made cheaply enough.
…
As for Marvell, the partnership with OLPC could lend more credibility to its Moby tablet, which is designed for educational uses like electronic textbooks. OLPC is also building Marvell’s chips into an upgraded form of its XO, known as the XO 1.75, later this year. “When we first met Nicholas, we were very moved by his leadership,” says Dai Weili, Marvell’s chief operating office. “We’ve got the cost structure, feature capability and scalability to support his vision for many years to come.”

Google Prediction API

This looks very cool.

The Prediction API enables access to Google’s machine learning algorithms to analyze your historic data and predict likely future outcomes. Upload your data to Google Storage for Developers, then use the Prediction API to make real-time decisions in your applications. The Prediction API implements supervised learning algorithms as a RESTful web service to let you leverage patterns in your data, providing more relevant information to your users. Run your predictions on Google’s infrastructure and scale effortlessly as your data grows in size and complexity.

Accessible from many platforms: Google App Engine, Apps Script (Google Spreadsheets), web & desktop apps, and command line.

The Prediction API supports CSV formatted training data, up to 100M in size. Numeric or unstructured text can be sent as input features, and discrete categories (up to a few hundred different ones) can be provided as output labels.

Uses:
Language identification
Customer sentiment analysis
Product recommendations & upsell opportunities
Diagnostics
Document and email classification
…

Teaching Through Tinkering

I wrote about the Tinkering School, Engineering camp previously. I am a strong believer in the value of helping kids (even adult kids – the few that haven’t resigned themselves to limited capacity to wonder since they now are grown up and not suppose to waste their time dreaming) explore their ideas and assisting them in making those ideas into reality. I think this is the best way to learn, not learning to pass a test, but learning to gain knowledge and accomplish things. Here is a nice 15 minute talk by the founder of the Tinkering School, Gever Tulley: “Turning Curriculum Design On Its Head: Engage First Then Look for Learning Within”

The format of the tinkering school is week long sessions where the kids stay overnight.

Some quotes: “we would use real tools and real materials and we would build real things, not model building, [but instead] actual building.” “create a meaningful experience and learning will follow”

Gever Tulley recently published: Fifty Dangerous Things (You Should Let Your Children Do).

Ironmaking at the National Museum of Science and Technology in Stockholm

Joakim Storck discusses pre–industrial Swedish and Japanese techniques for iron and sword making from a museum demonstration at Tekniska museet. Ironmaking at the National museum of science and technology, Stockholm 2005

Bricklaying is a messy story. The mortar consists of clay, sand and horse manure (if available), mixed with water to a fairly loose batter. The best finish is obtained if you work with your hands as the mortar is placed on, and smears with water so that the surface becomes smooth and fine. When then furnace is ready, it is dried through slow heating by wood without blasting, until the moist has been driven out of the mud. At this stage, heating should be quite cautious in order to avoid cracking.

Then, on Friday September 9, we went again with a fully loaded trailer from Dalarna in the direction of Stockholm. More than a few people were probably turning their heads when we passed, because the trailer was dominated by a large bellow — our newly built two chamber bellow with an estimated bladder capacity of up to 800 litres per minute. In addition, we brought fire wood, iron rods, pliers, some stumps and other stuff needed for the furnace operation.
…
We made one run each on Saturday and Sunday. Each time we charged a total of about 10kg ore added in amounts of about 1kg every 20 minute. For each charge, we added about twice the amount of charcoal. Discharging of the loupe was scheduled for two o’clock, and by that time a fairly large crowd had gathered to see the show. This time we managed to get the loupe out of the furnace without too much trouble. Worse was that the process took longer than expected, but the crowd seemed to be patient and people stayed around until the end.

Webcast on Finding the Missing Memristor

Very interesting lecture on finding the missing memristor by R. Stanley Williams. From our post in 2008:

How We Found the Missing Memristor By R. Stanley Williams:

For nearly 150 years, the known fundamental passive circuit elements were limited to the capacitor (discovered in 1745), the resistor (1827), and the inductor (1831). Then, in a brilliant but underappreciated 1971 paper, Leon Chua, a professor of electrical engineering at the University of California, Berkeley, predicted the existence of a fourth fundamental device, which he called a memristor.

Google Social Circle Results

I haven’t paid much attention to this before: Google showing results based on your social network. In genera,l I think Google is doing great stuff. Their approach to profiles, buzz, search wiki, and social stuff in general however, I find poor (extremely poorly when I see how well they do so much else). They are too borg-like in their insistence you do things exactly as proscribed by them. They opt you in far to often, in the way they want – completely and with few, if any, options. They don’t provide good tools to let you manage your profile and connections. The Google profile itself is extremely lame.

It is very silly that they don’t let you create personas you want to use and let you use them as you want. They force you to use the Google account you are logged in as to access Google services as the profile used in Google searches. I don’t want or need the 2 tied together. And I would much prefer a way to switch between my personas by search (or buzz [though I dropped it because it was so inflexible] or whatever). I know which persona I want for a specific search. This seems like a very obvious thing lots of people would want to do. Google’s whole monolithic, one-very-rigid-size-fits-all social solutions don’t allow this. It is a fundamental flaw. Without fixing the flexibility of social services from Google I see them having trouble succeeding in that area. On the other hand much else of what they do is fantastic.

The way their search wiki stuff works is very similar: inflexible. They seem focused on do it how we want which is not, I think, the way most people want. Their social solutions are very all or nothing. They want people to behave how Google wants. For that reason after short attempts to try Google’s social efforts I give up. I keep hoping they will become more flexible and user friendly but keep being disappointed.

This is the network of connections Google uses to identify relevant social search results. It is based on a combination of the following:

Direct connections from your Google chat buddies and contacts (5)
Direct connections from links listed on your Google profile (0) such as Twitter and FriendFeed
Secondary connections (252) that are publicly associated with your direct connections
In addition to web pages from your social circle, posts from your Google Reader subscriptions may also appear in your social search results.

Google Social Search is a feature designed to help you discover relevant publicly-accessible content from your social circle, a set of online friends and contacts. The idea is that content from your friends and social contacts is often more relevant to you than content from strangers. For example, a movie review from an expert is useful, but a movie review from your best friend can be even better.

See video on Google social search:
Continue reading →

Next steps for Google’s Experimental Fiber Network

Think big with a gig: Google’s experimental fiber

Universal, ultra high-speed Internet access will make all this and more possible. We’ve urged the FCC to look at new and creative ways to get there in its National Broadband Plan – and today we’re announcing an experiment of our own.
…
We’re planning to build and test ultra high-speed broadband networks in a small number of trial locations across the United States. We’ll deliver Internet speeds more than 100 times faster than what most Americans have access to today with 1 gigabit per second, fiber-to-the-home connections. We plan to offer service at a competitive price to at least 50,000 and potentially up to 500,000 people.

Next steps for our experimental fiber network

So what’s next? Over the coming months, we’ll be reviewing the responses to determine where to build. As we narrow down our choices, we’ll be conducting site visits, meeting with local officials and consulting with third-party organizations. Based on a rigorous review of the data, we will announce our target community or communities by the end of the year.

Of course, we’re not going to be able to build in every interested community — our plan is to reach a total of at least 50,000 and potentially up to 500,000 people with this experiment. Wherever we decide to build, we hope to learn lessons that will help improve Internet access everywhere.

This is another great idea from Google. Not only to push forward the much poorer internet connectivity those in the USA have than other countries but it will hopefully lead to some real engineering breakthroughs. And it is a smart move to increase Google’s potential income – a better internet experience (for users) will likely help Google quite a bit.

Basketball Padding

Basketball used to be considered a non-contact sport. Now more and more college and pro players are wearing padding. March Madness, this year with more padding

Plumlee and dozens of other college basketball players wear compression shirts and shorts dotted with foam and plastic shock-absorbing pads under their uniforms. There are also padded sleeves for the elbows and knees. In the past few years players have started to wear this layer of protective gear meant to feel like a second-skin in a sport that has bigger, faster and stronger athletes than ever.
…
“Some of the things you’ll see like these products, a lot of them tend to be more fads that come and go,” he said. “But anything that comes down over the edge of a bony prominence, or on the knee, makes sense. For the ribs — there’s cartilage that is a natural shock absorber so I don’t know how truly affective that piece might be.”

Purchase: McDavid knee and elbow pads – McDavid Hex Power Shooter Arm Sleeve – Hexpad Thudd with Extended Thigh

Autonomous Underwater Robot Decides on Experiment Options

Ocean robot plans experiments

the autonomous underwater vehicle (AUV) used a piece of software called “T rex”, which operates in a similar way to the software used to control Nasa’s Mars Exploration Rovers – helping them to avoid obstacles on the surface of the Red Planet.

One main difference between the two pieces of software is that for the Mars rovers, the software ran in the control centre on Earth. With this marine vehicle, it runs onboard the robotic vehicle.
…
“We tell it, ‘here’s the range of tasks that we want you to perform’, and it goes off and assesses what is happening in the ocean, making decisions about how much of the range it will cover to get back the data we want.”

Researchers at MBARI used the Gulper AUV to monitor potentially harmful algal blooms.

Kim Fulton-Bennett from MBARI explained: “We used to send out a ship for a full day every few weeks to manually take these measurements. Now we just take the AUV outside the harbour and send it on its way.

“About 24 hours later, it comes back, we hoist it on board, and download the data.”