t e c h i e - t a l k s ::: technology at its best

Intel says parallel software is more important for many-core CPUs like “Larrabee”
Multi-core processors have been in the consumer market for several years now. However, despite having access to CPUs with two, three, four, and more cores, there are still relatively few applications available that can take advantage of multiple cores. Intel is hoping to change that and is urging developers of software to think parallel.

Intel director and chief evangelist for software development products talked about thinking parallel in a keynote speech he delivered at the SD West conference recently. James Reinders said, “One of the phrases I’ve used in some talks is, it’s time for us as software developers to really figure out how to think parallel.” He also says that the developer who doesn’t think parallel will see their career options limited.

Reinders gave the attendees eight rules for thinking parallel from a paper he published in 2007 reports ComputerWorld. The eight rules include — Think parallel; program using abstraction; program tasks, not threads; design with the option of turning off concurrency; avoid locks when possible; use tools and libraries designed to help with concurrency; use scalable memory; and design to scale through increased workloads.

He says that after half a decade of shipping multi-core CPUs, Intel is still struggling with how to use the available cores. The chipmaker is under increasing pressure from NVIDIA who is leveraging a network of developers to program parallel applications to run on its family of GPUs. NVIDIA and Intel are embroiled in a battle to determine if the GPU or CPU will be the heart of future computer systems.

Programming for processors with 16 or 32 cores takes a different approach according to Reinders. He said, “It’s very important to make sure, if at all possible, that your program can run in a single thread with concurrency off. You shouldn’t design your program so it has to have parallelism. It makes it much more difficult to debug.”

Reinders talked about the Intel Parallel Studio tool kit in the speech, a tool kit for developing parallel applications in C/C++, which is currently in its beta release. Reinders added, “The idea here [with] this project was to add parallelism support to [Microsoft’s] Visual Studio in a big way.”

Intel says that it plans to offer the parallel development kit to Linux programmers this year or early next year. The CPU Reinders is talking about when he says many-core is the Larrabee processor. Intel provided some details on Larrabee in August of 2008.

One of the key features of Larrabee is that it will be the heart of a line of discrete graphics cards, a market Intel has not participated in. Larrabee is said to contain ten of more cores inside the discrete package. If Larrabee comes to be in the form Intel talked about last year it will be competing directly against NVIDIA and ATI in the discrete graphics market.

NVIDIA is also rumored to be eyeing an entry into the x86 market as well. Larrabee will be programmable in the C/C++ languages, just as NVIDIA’s GPUs are via the firms CUDA architecture.

Source:  Dailytech.com

A new paradigm for using web applications has emerged. The landing page looks similar to Google, with a simple layout and that familiar search box. But what lies beneath this new search engine is something powerful and breath-taking. Meet Wolfram Alpha - Brings search engines to a whole new level. The algorithm sounds extremely smart to a certain extent, it sounds just as scary too!

Unlike your normal search via keywords, Wolfram Alpha uses what they call a Knowledge Engine to churn out search results that are not keyword-based but rather, question-based strings. For example, instead of searching “Google+birth date”, a typical search on this search engine would be “When Google was born?” The algorithm is a combination of natural language and powerful mathematics.  You can search for something like “What is the 99th digit of Pi” and it’ll give you the right answers!

But science is just one of the domains it knows about — it also knows about technology, geography, weather, cooking, business, travel, people, music, and more.

It also has a natural language interface for asking it questions. This interface allows you to ask questions in plain language, or even in various forms of abbreviated notation, and then provides detailed answers.

The algorithm is written by Wolfram, who at the age of 20, got his first Ph.D. in theoretical physics. In 1988 he launched Mathematica, powerful computational software that has become the gold standard in its field. In 2002, Wolfram produced a 1,280-page tome, A New Kind of Science, based on a decade of exploration in cellular automata and complex systems.

This site is a must read. It describes the functions and capabilities of this new search engine. However, Wolfram Alpha will not be like HAL9000 or even a real-life Skynet. It will not be able to recognize feelings or even tones in a sentence. However, I don’t see that such an implementation (at least the latter) would be far from the near future. However scary it may seem, it is far from being able to cause any harm.

I took some time out of my busy schedule to attend a talk by Mario Behling. The talk was organized by the people from LinuxNus. A short intro of Mario taken from the LinuxNus website:

Mario Behling is a FOSS advocate known in the international (Free and Open Source Software) FOSS community through his work for FOSS Bridge, Freifunk and LXDE. Mario, originally from Berlin, has lived in different countries in Asia, Europe and in Australia. He works with business and educational projects like the OLPC project in Afghanistan or FOSS Bridge in Vietnam, where he advises companies on open source business and internationalization strategies and organizes matchmaking events for SMEs. Mario was elected as the president of the LXDE Foundation in 2008. As an active community manager he promotes the LXDE community and coordinates the work of projects like the LXDE Translation Project. As part of his engagement in the free wireless freifunk community he maintains freifunk websites like the wiki (wiki.freifunk.net) and the global free wireless newswire (global.freifunk.net).

The talk was motivational and very inspirational. It is really amazing how people are able to take time out of their daily lives to contribute to the open source community as developers and even non-developer positions such as open source event organizers. I felt rather inspired, to the extent of wanting to develop a new distro (a long unfulfilled dream). Thus, I did a rather short thinking session on my walk home.

There’s a Linux distribution for almost any and every situation you can think of. From full-fledge application-heavy developer’s distros to a small operating system that can fit in your thumb drive. But I realized (maybe I might be wrong) that there isn’t actually a Linux operating system ditribution streamlined for “smart-home” embedded systems (then again, maybe there might not be a need for one). Hence, I began listing some functional and non-functional requirements for such an operating system.

Functional Requirements:

1. Able to communicate with any household interface through wired/wireless interfaces.
2. Handle a lot of unpredictive events (event triggered applications).
3. Nice touch-screen GUI.
4. Possible NAS-to-OS interface for storage/streaming of multimedia.
5. Ability to detect and prevent intrusions.
6. Will be connected to almost everything in the house (Fridge, kettle, TV, printers, multimedia box, consoles, etc).
7. Robust with very small context switching delays of tasks.
8. Small memory footprint. Lightweight and require < 4Gb space.
9. Good handle of I/O operations.
10. …..

Non-functional requirements:

1. Processor with excellent I/O capabilities (x86 compatible?).
2. Ethernet and Wifi (802.11 b/g/n) connections.
3. Touch screen (haptic).
4. Compatible with some smart home interface and standards (if any?).
5. …
6. …

Well, these are just the lists of requirements that came from the top of my mind while walking back home. It’ll definitely need refinements and research. Just imagine being able to control the temperature of your fridge from your office, stock-check your foodstuffs from school, record your favourite TV show from overseas, monitor your house remotely, switch on/off the lights or any connected electrical appliances for that matter from anywhere, VoIP from a remote location via your house phone and lots more! A geek heaven I must say!

Anyone keen to take on the challenge? haha.. Well, maybe after I hand in my Final Year Project. boo!

The Linux Foundation insists it is equipped to fight Microsoft if the software giant’s lawsuit against TomTom impacts the open source Linux kernel.

In his blog, Executive Director Jim Zemlin advised concerned parties to “calm down” in light of statements by Microsoft’s deputy general counsel that it is targeting TomTom’s GPS mapping software and not Linux.

Yet Zemlin was quick to reassure the community that the foundation and Open Innovation Network has the funds necessary to defend Linux if necessary. TomTom’s GPS mapping software incorporates the Linux kernel. Below is an excerpt of Zemlin’s blog:

Calm Down

Right now the Microsoft claim against Tom Tom is a private dispute between those two entities concerning GPS mapping software. We do not feel assumptions should be made about the scope or facts of this case and its inclusion, if any, of Linux-related technology. Any patent litigator will tell you that the path between asserting a claim under a patent and an actual, final determination that the patent is (1)valid and (2) that the claims of the patent are actually infringed is an extremely long road. If this case is in any way directed at Linux (in fact, Horacio Gutierrez, Microsoft’s corporate vice president and deputy general counsel of Intellectual Property and Licensing, has specifically stated that it isn’t), the Linux ecosystem has enormously sophisticated resources available to assist in the defense of any claim that is made against Linux.

Hope for the Best

It is our sincere hope that Microsoft will realize that cases like these only burden the software industry and do not serve their customers’ best interests. Instead of litigating, we believe customers prefer software companies to focus on building innovative products.

Plan for the Worst

The Linux Foundation is working closely with our partner the Open Invention Network, and our members, and is well prepared for any claims against Linux. We have great confidence in the foundation they have laid. Unfortunately, claims like these are a by-product of our business and legal system today. For now, we are closely watching the situation and will remain ready to mount a Linux’s defense, should the need arise.

And it just might arise. In her blog,

Mary Jo Foley cites an online report by TechFlash’s Todd Bishop, which maintains that Microsoft in its lawsuit alleges that TomTom’s products violate eight of its patents, including three related to TomTom’s implementation of the Linux kernel.

Microsoft’s Horacio Gutierrez, Microsoft’s corporate vice president and deputy general counsel of Intellectual Property and Licensing, said the case targets GPS mapping software — not Linux per se — but we’ll have to wait and see how the patent claims pan out.

I can think of at least two reasons why Microsoft chose to take on TomTom: TomTom’s new personal navigation devices carry a Go Live and Live Services product naming — the same branding used by Microsoft for its next gen web services. And those devices just hit the U.S. market last month.

Source: zdnet.com

BRING IT ON!

Microsoft stunned its investors on Thursday, announcing the first broad layoffs in its history and offering a pessimistic forecast for the second half of its fiscal year.

Rather than issuing its second-quarter results in the customary fashion after the market closed, Microsoft rushed out the news Thursday morning that it will lay off up to 5,000 of its 94,000 employees over the next 18 months, including 1,400 people Thursday. The layoffs span across research, sales, finance and technology roles, the company said.

“We will continue to manage expenses and invest in long-term opportunities to deliver value to customers and shareholders, and we will emerge an even stronger industry leader than we are today,” said Steven A. Ballmer, Microsoft’s chief executive.

Microsoft’s shares dropped $1.50, or 8 percent, to $17.88 in early trading.

For its second quarter, Microsoft posted net income of $4.17 billion – a figure 11 percent lower the $4.71 billion reported in the comparable period last year. Microsoft’s revenue for the quarter rose 2 percent year-over-year to $16.63 billion.

Microsoft’s earnings of 47 cents during the quarter missed the forecast from Thomson Reuters by 2 cents.

The direct impact of falling personal-computer sales, which roiled Intel last week, were evident in Microsoft’s results, as sales of its PC operating-system software dove 8 percent to $3.98 billion from $4.33 billion last year.

Blaming market uncertainty, Microsoft declined to issue a revenue or earnings forecast for the rest of its fiscal year.

“We are planning for economic uncertainty to continue through the remainder of the fiscal year, almost certainly leading to lower revenue and earnings for the second half relative to the previous year,” said Chris Liddell, the company’s chief financial officer.

Source: The New York Times

Silicon Valley chipmaker Advanced Micro Devices will eliminate 1,100 jobs - roughly 9 percent of its workforce - in the first quarter, according to a report on Marketwatch. The company also said it will suspend its 401(k) match for employees and that executive chairman Hector Ruiz and CEO Dirk Meyer each will temporarily take a 20 percent base salary cut.

There will be smaller percentage salary cuts for other employees, as well - 15% for vice presidents and other execs; 10% for nonovertime-eligible employees; and 5% for all overtime-eligible employees. The company also plans to take a $622 million goodwill impairment charge related to its 2006 acquisition of ATI Technologies acquisition and will also post a $62 million charge related to the purchase.

In a statement, the company said:

As a result of the continuing global economic downturn, we have determined that we need to take difficult but prudent actions designed to reduce our costs. Beginning in February, we are undertaking several steps to lower costs, including temporarily reducing employee base pay and suspending some benefits programs.

Last year, the company said it plans to spin off its manufacturing business into a joint venture largely funded by Abu Dhabi investment companies, a move that would bring in the billions of dollars it was struggling to find to construct state-of-the-art chip plants needed to build the smaller, faster and more energy-efficient chips that compete with Intel’s.

The rough economy is hitting the chip industry hard. Yesterday, Intel reported a horrible fourth quarter - one of the worst since 2000. CEO Paul Otellini told analysts on a conference call that it was “only the second time in 20 years where revenues declined in the fourth quarter from third quarter.” Larry Dignan wrote in his post yesterday:

In a nutshell, Intel was whacked by a demand shutdown that led to high inventory and factory underutilization. Pricing wasn’t an issue since rivals like AMD were also whacked.

Still, the short-term doom-and-gloom future wasn’t stopping AMD from talking about how it’s positioned for the rough economic road ahead. AMD’s global VP of advanced marketing Pat Moorhead sat down with ToyBox blogger Andrew Nusca at last week’s Consumer Electronics Show to discuss AMD’s latest products and the plan for 2009. Moorhead said, “AMD is positioned well for this economy,” referencing the company’s new product focus with regard to the current recession. According to Moorhead, AMD is targeting what he called the “sweet spot” of consumer: the $999 price point. With it, he’s offering an “enthusiast dream machine” in the form of the $699 HP dv2.

Source: zdnet.com

Medfield will feature memory controller, graphics, and I/O on single die

A new roadmap has cropped up that shows the future of Intel’s Atom processors that power the majority of netbook computers on the market today. The Atom has proven to be very successful for Intel and future versions will offer even more power and performance.

CNET News reports that the new roadmap shows a new Medfield Atom will debut in 2010. The processor will be built on the 32nm process, where as the current generation Atom parts are built on the 45nm process.

Before 2010 when Medfield pops up, Intel will be releasing Atom processors code named Pineview in 2009. DailyTech has covered the Pineview Atom processors before. The biggest difference between Pineview and today’s Atom CPU will be the integration of a GPU core and memory controller onto the processor die.

In 2010, Medfield will integrate onto the die the processor, memory controller, multimedia functions, and I/O onto a single chip. A graphics core would be integrated as well. With this change, the netbook will gain the ability to handle HD video and move from a basic web surfing device into a full function entertainment device.

The information on the future Atom parts comes from a report created by UBS Securities. CNET News quotes the UBS Securities report saying, “[netbooks will evolve] from basic web page consumption to multimedia consumption including high-definition (HD) video.”

Another bit of good news to look forward to is that battery life will increase for systems using the new Atom parts. Even with new features like touchscreen and WAN access the run time of netbooks will move from two to three hours up to near five hours according to the report.

Dual-core Atom parts are expected to be offered with the Medfield just as the current generation Atom parts feature dual-core variants.

Source: Dailytech.com

Whether the ailing economy is real or imagined, you might be looking for ways to shed some pounds off your bloated IT budget. You should seriously consider Linux and Open Source software to put your 2009 financial appetite on a diet. Here are 10 reasons to choose Linux in a bad economy and none of them have to do with the fact that Linux is free.

10. Value - Linux has value to a worldwide group of consumers. It runs most of the world’s mission critical websites and is the platform of choice for server virtualization.

9. Less Hardware Overhead - You can still run world class websites, applications, and services on a machine that is outdated by today’s standards. And you have the added bonus of being able to outrun those applications compared to their Windows counterparts.

8. Active Development - Linux is actively developed meaning that new hardware drivers are available as soon as a new device hits the market and improvements are being made on a daily basis to the kernel and supporting code. No need to wait two or three years for the next belated and bloated version from the other guys.

7. Choice - Linux gives you a choice to do things differently and better in your home, office, or data center. Having a choice is good for consumers because it means that vendors and programmers are trying to get your attention by creating quality products for you to use. Competition creates better products and services which is a boon to you and your budget.

6. Multiple Distributions - I’ve heard this one used as a negative part of a campaign by Microsoft claiming that there are too many Linux distributions. Balderdash! Having a system that meets any need from cell phones and wristwatches to supercomputers is just what we need to solve the problems at hand. Having only one distribution is like needing to build a house but finding only a screwdriver in your toolbox.

5. Open Source Model - The Open Source Development model helps everyone in the Linux community from end users to other developers all the way up to C-level executives. The way in which this model helps is that all development is open and source code is available for all to see and improve upon. You can create applications and alter the Operating System itself for any purpose.

4. Available Development Community - Want to contact the developers who created a particular application, protocol, or service? You can. You can submit bug reports, email them directly, and in some cases speak directly to them on the phone. You can even submit your own code that will be included in a major distribution. You have real input to Linux and Open Source software.

3. Unix Stability - Linux is based on Unix and shares its multi-tasking, multi-user, and stable kernel and filesystem structures. For mission-critical environments, you need this kind of stability. Stability means not only that you don’t have to reboot the machine for software installations, driver updates, or even network changes but also that you’ll enjoy uptimes measured in years not days.

2. Compatibility - The old argument was that Linux wasn’t compatible with Windows but through the magic of Samba (File and Printer sharing), Wine, and Cygwin; Linux and Windows are very interoperable and can share files, applications, and services.

1. Commercial Support - Another old complaint from the Redmond camp was that Linux is supported by a bunch of amateurs and part-timers. Red Hat, Novell, Xandros, IBM, HP, Dell, Canonical, and others have thrown their significant financial and human resources behind Linux. Linux has commercial support–even from Microsoft.

In a downturned economy, you have to be creative and look for innovative solutions that are cost-effective, supportable, and protective of your business. By choosing Linux, you have all those options at your disposal.

Source: Daniweb.com

It seems like mainstream multicore CPUs are going their own way this time. An increasing number of cores in today’s CPUs leads to a decreasing performance in supercomputers. The major obstacle is not a brick wall, rather it is what engineers call - memory wall.

1. Memory speeds aren’t keeping up with CPU speeds.
2. With multicore CPUs, alot of bandwidth is wasted on keeping caches coherent.

Therefore, after a certain number of cores in a single-die CPU, a huge performance loss is seen. Taken from IEEE:

Trouble Ahead: More cores per chip will slow some programs [red] unless there’s a big boost in memory bandwidth [yellow].

With no other way to improve the performance of processors further, chip makers have staked their future on putting more and more processor cores on the same chip. Engineers at Sandia National Laboratories, in New Mexico, have simulated future high-performance computers containing the 8-core, 16?core, and 32-core microprocessors that chip makers say are the future of the industry. The results are distressing. Because of limited memory bandwidth and memory-management schemes that are poorly suited to supercomputers, the performance of these machines would level off or even decline with more cores. The performance is especially bad for informatics applications—data-intensive programs that are increasingly crucial to the labs’ national security function.

High-performance computing has historically focused on solving differential equations describing physical systems, such as Earth’s atmosphere or a hydrogen bomb’s fission trigger. These systems lend themselves to being divided up into grids, so the physical system can, to a degree, be mapped to the physical location of processors or processor cores, thus minimizing delays in moving data.

But an increasing number of important science and engineering problems—not to mention national security problems—are of a different sort. These fall under the general category of informatics and include calculating what happens to a transportation network during a natural disaster and searching for patterns that predict terrorist attacks or nuclear proliferation failures. These operations often require sifting through enormous databases of information.

For informatics, more cores doesn’t mean better performance [see red line in “Trouble Ahead”], according to Sandia’s simulation. “After about 8 cores, there’s no improvement,” says James Peery, director of computation, computers, information, and mathematics at Sandia. “At 16 cores, it looks like 2.” Over the past year, the Sandia team has discussed the results widely with chip makers, supercomputer designers, and users of high-performance computers. Unless computer architects find a solution, Peery and others expect that supercomputer programmers will either turn off the extra cores or use them for something ancillary to the main problem.

At the heart of the trouble is the so-called memory wall—the growing disparity between how fast a CPU can operate on data and how fast it can get the data it needs. Although the number of cores per processor is increasing, the number of connections from the chip to the rest of the computer is not. So keeping all the cores fed with data is a problem. In informatics applications, the problem is worse, explains Richard C. Murphy, a senior member of the technical staff at Sandia, because there is no physical relationship between what a processor may be working on and where the next set of data it needs may reside. Instead of being in the cache of the core next door, the data may be on a DRAM chip in a rack 20 meters away and need to leave the chip, pass through one or more routers and optical fibers, and find its way onto the processor.

In an effort to get things back on track, this year the U.S. Department of Energy formed the Institute for Advanced Architectures and Algorithms. Located at Sandia and at Oak Ridge National Laboratory, in Tennessee, the institute’s work will be to figure out what high-performance computer architectures will be needed five to 10 years from now and help steer the industry in that direction.

“The key to solving this bottleneck is tighter, and maybe smarter, integration of memory and processors,” says Peery. For its part, Sandia is exploring the impact of stacking memory chips atop processors to improve memory bandwidth.

Source: Moore, Samuel K., Multicore is Bad for Supercomputers, IEEE, Nov 2008

According to Erricson’s chief technology officer, John Cunliffe, mobile broadband could hit an unimaginably high speed of 42 Mbps by early next year. The current highest bandwidth in Singapore is 7.2 Mbps on a HSDPA network (Starhub, Singtel & M1).

The increase in speed is achieved by having better better antenna design and coding. Quoted from PC Pro,

“If you look at what’s actually happening, the fastest service in the UK at the moment is 7.2Mb/sec, but we can easily see how we get to 14Mb/sec by improving the codes using the modulation. “

“After that there are two steps that can boost speed. We go to the modulation that is 64 QAM… that’s 64 combinations of information in the same slot as one piece of information. In other words it’s massively spectrally effecient. We think that’s going to give us 21Mb/sec.”“But then there’s another technique called MIMO, multiple in multiple out. What this means is you have multiple radios on a device, this is like Wi-Fi uses with the N standard.

With MIMO we can go from 14Mb/sec to 28Mb/sec. They’re two paralell paths, but we can then combine them to get 42Mb/sec. Essentially this is happening through 2009.”

Cunliffe acknowledges that while the technology to boost speeds exists, it may take another year or so for the dongles themselves to catch up. However, he doesn’t think speeds will stop at 42Mb/sec.

“We think even more in the future, we can probably squeeze that to 80Mb/sec, and that’s before we even get to Long Term Evolution.”

Source: PCpro.co.uk