As IT job opportunities in the United States and Europe start to contract -- a trend that predates the current financial meltdown but may accelerate because of it -- perhaps it's time to look abroad, where there may, in fact, be more growth and better opportunities to advance your career.

For example, the International Monetary Fund now predicts that the U.S. economy will barely grow -- at a 0.1 percent rate -- in 2009, while European economic growth will range from a slight contraction (-0.2 percent) in Italy to slight growth (0.2 percent) in France. Canada will be the superstar of the major developed economies, predicted to grow 1.2 percent.

By comparison, the IMF expects China to grow at about 8 percent, with India at about 7 percent, and Russia about 6 percent -- despite the financial crisis. The world at large should grow about 3 percent.

[ Use InfoWorld's interactive map to learn about 12 hot cities and 6 regions you should consider for career-boosting tech jobs abroad. ]

Tech jobs overseas are no longer just the scut work of heads-down programming. As foreign, U.S., and global firms have set up shop throughout the world, they've increased the demand for a wide range of tech talent in those locations. Foreign companies are particularly looking for IT professionals with business fluency, and such experience is more common in the United States than in most places. The combination of industry-specific skills and knowledge of American markets is an invaluable asset that outsource providers from countries like China, India, and Russia lack and will pay a premium for.

And the experience you'll gain from working overseas will make you even more valuable. According to Rob McGovern, CEO of JobFox, an international employment agency for IT, in today's global economy, people who truly understand how to do business globally are a minority. "IT is going global. The IT profession is going global. Developing product for markets all over the world is something you have to learn how to do. Overseas work is a huge enhancer for IT professionals," he says.

So where should you move to accelerate your tech career? InfoWorld interviewed outsource suppliers and industry analysts from around the world and found 12 hot cities and six promising regions, as well as what it takes to make the move. Many are in the Far East of Asia and in Eastern Europe, but Latin America, the Middle East, and -- closer to home -- Canada are all strong possibilities as well, for at least some tech skills.

The most popular types of tech jobs outside the United States vary considerably, and emerging IT centers are themselves trying to diversify their own areas of expertise. But as a gross generalization, product support and business process development positions are more likely to be in India than in Indianapolis; embedded software development positions are more likely to be in China than Cincinnati.

[ Story? continued below the map. ]

The paths to working overseas So how realistic is it really to move overseas for work? The answer varies based on the country and, of course, your personal circumstances. Family considerations -- such as finding a job for your spouse and a school for your children -- can make an overseas move much harder for a family than for a single person. In terms of the basic process, however, there are three routes to getting a job overseas.

The first is to get a work visa in the destination country, the equivalent of the H-1B program in the United States. This typically requires that the employer sponsor you and go through a process proving you are not taking a position a local could fill.

The second is to get a work-rotation visa in the destination country, the equivalent of the L-1 program in the States. This type of visa lets companies rotate employees among their offices in various countries. It's often used for executives to help them gain experience across different corporate units but can be applied to other positions as well. Global consultancies, federal agencies, and multinationals are the typical venues for such positions.

The third is to use any dual nationality you may hold, such as from being the spouse or child of a foreign national, to seek work in that other country. After all, as a citizen of that country, you have the same employment rights as any other national. (The fact that you are also a U.S. citizen doesn't matter, at least in countries that allow dual citizenship.)

The fourth is to set up your own company in the United States and be a consultant overseas.

Some locales, like Costa Rica, actually make it easier for foreigners to come in and start a company rather than come in as an employee who might be taking a job away from a local.

The top regions and cities to explore for overseas tech jobs Based on dozens of interviews, InfoWorld has come up with the following regions and cities worth exploring if you want to offshore yourself:

Regions: * Asia * Canada * China * Europe * India * Latin America

Cities: * Amsterdam, the Netherlands * Bangalore, India * Dubai, United Arab Emirates * Dublin, Ireland * Hong Kong, China * Kiev, Ukraine * New Delhi, India * Paris, France * San Jos?, Costa Rica * S?o Paulo, Brazil * Shanghai, China * Tel Aviv, Israel

Microsoft Office Labs released Tuesday its Touchless SDK for developers to experiment with multi-touch technology.

With the SDK, developers can build multi-touch applications that use a color marker and webcam for input. Utilizing the webcam, applications can track the location and size of the color marker to enable multi-touch functionality without touching a surface, according to Microsoft.

The SDK can be used to build games or applications that could, for example, browse through media, said project developer Mike Wasserman. A demonstration of the technology can be viewed here.

"I could see this as a great way to say, browse media like pictures, music, and video on your computer," Wasserman said. A user could, for example, could use hands to manipulate an image on a screen and give it annotations.

With the SDK, which is available on Microsoft's CodePlex site, developers can view, use, and contribute to the prototype code. The SDK leverages Microsoft's DirectShow API for media streaming.

"Once you have a webcam installed, you can use our SDK to create these applications," Wasserman said.

The Power 570, IBM's top-selling midrange server, is now available with a 5GHz Power6 processor, which was previously available only in IBM's high-end Power 595 system, as IBM targets customers who want to use its hardware for virtualization and server consolidation. The manufacturing yields for the 5GHz chip have been good enough that IBM can offer it in higher-volume systems, said Scott Handy, IBM vice president of worldwide marketing and strategy.

IBM also doubled the maximum density for the Power 570, so customers can put up to eight 4.2GHz processors in each server node, or up to four 5GHz processors. The nodes can be stacked four high for a total of 32 4.2GHz processors in a single box.

The enhancements boost the performance customers can get from each square foot of server space and from each watt of power, making the 570 well-suited for virtualization and consolidation, Handy said.

IBM is also testing a new capability for its PowerVM virtualization software that allows administrators to share virtual system memory between partitions, in the same way they can share virtual processors today. Called Active Memory Sharing, the technology is in closed beta and due for general availability in the first half of next year.

Also planned is a new version of Active Energy Manager that lets administrators cap the energy being used by a pool of servers. Due later this quarter, the software will throttle down processor and fan speeds if an application tries to exceed a certain power threshold. The software will be a plug-in for IBM Systems Director . The portion for monitoring power use is free, but the tool for throttling down consumption will cost about $275 per server for the Power 570, an IBM spokesman said.

IBM hopes the improvements will help keep it on top of the Unix server market. IBM regained its lead in Unix server revenue in the second quarter, growing its share 5 percentage points from a year earlier, according to IDC. Sun Microsystems, in second place, dropped 5.6 percent percentage points over the same period, while third-place Hewlett-Packard gained 1 point, IDC said.

"Ever since their resurgence in the Unix market, IBM has been pushing performance up and prices down when they really don't have to," said Dan Olds, principal analyst at Gabriel Consulting Group in Beaverton, Oregon. "Right now they have the performance lead, and they are still pushing the speeds and feeds higher. They're keeping constant pressure on Sun and HP."

Among the other IBM Unix news Tuesday:

* A new technology for ensuring high availability for applications, called hot node and repair, will be available later in the quarter for the Power 570 and 595 systems. If a server indicates it is about to fail, the software allows an administrator to shift a workload to a different server while the first server is repaired, and then move back again without losing transactions, according to IBM.* IBM doubled the processor cores available in its System i servers, which run its i OS, to bring them on par with its AIX and Linux offerings. The System i 550, for example, now comes with up to eight processor cores, up from four. IBM said the move completes an effort to offer a common hardware platform for its servers running Linux, AIX or i5/OS.* A new server for mid-size companies, the Power 560 Express, is due on Nov. 21. It uses a 3.6Ghz Power6 processor, comes in four-, eight- and 16-node configurations, and packs a hefty 384GB of memory. It's designed for companies looking to run multiple applications on a virtualized system. It will be offered with Linux, AIX or i. Pricing wasn't announced.

--IBM said its JS12 Express blade server can now be pre-installed with its i software and attached to its DS3200 storage gear to create a low-cost SAN.

-- Handy said almost two-thirds of IBM's Unix servers shipped in the second quarter included its PowerVM virtualization software, up from 21 percent in the second quarter the year before.

Nippon Telegraph and Telephone Corporation (NTT), Japan's largest telecommunications company, has formed a "wide-ranging strategic partnership" with EnterpriseDB that includes an undisclosed financial investment in the open source database maker.

The NTT announcement follows news earlier this year that IBM had become an investor in EnterpriseDB.

[ Track the latest trends in open source with InfoWorld's Open Sources blog. ]

"Many startup [database] vendors have one validating investment from a big company. It's pretty unusual, however, to have two," said Curt Monash, president of Monash Research.

NTT and EnterpriseDB will work together to improve the PostgreSQL database technology at the heart of the vendor's product line, focusing on high availability and scalability, and to promote adoption of PostgreSQL both within and outside NTT, according to a statement.

In an interview, EnterpriseDB CEO Ed Boyajian declined to reveal the size of NTT's investment, which he said had been in the works for "quite some time."

PostgreSQL, which competes with other open source databases like Sun's MySQL, is "useful for a broad range of applications," Monash said. "There are many reasons not to port from your existing brand-name [database management system], but if you're looking at a new application, you should look very hard at PostgreSQL, rather than building on a closed-source DBMS."

For its part, EnterpriseDB has positioned itself as a ready alternative to the dominant proprietary vendors, releasing an update to its Advanced Server product last year that made it easier to run applications written for Oracle on its own system.

Boyajian said he expected NTT, which could not immediately be reached for comment Tuesday, to use EnterpriseDB for new projects and to replace other platforms.

Dell is putting together hardware and software for backup and recovery in one system, with integrated software, for small and medium-size businesses.

The need for large-scale backups is growing as organizations produce more data and face regulatory requirements to hold onto it longer, analysts say. With its PowerVault DL2000 platform, which will go on sale later this month, Dell aims to make it easier for SMBs and branch offices with limited IT expertise to centralize and automate their backups. It can make those backups as much as 55 percent faster for disk storage and 77 percent faster for tape, Dell claims.

[ Get the latest on storage developments with InfoWorld's Storage Adviser blog and Storage Report newsletter. ]

The platform consists of the DL2000 -- a Windows-based x86 server running either a CommVault or a Symantec software stack -- and a Dell PowerVault MD1000 disk array. As an option, customers can have a PowerVault TL2000, TL4000, or ML6000 Tape Library added on for long-term archiving. The whole system will be put together in the factory and shipped as a turnkey solution, said Sanjeet Singh, a senior product marketing manager. The company designed the system as an alternative to buying several components and integrating them in-house, and claims an IT generalist can get the system started in less than half an hour.

That claim should prove true, said analyst Arun Taneja of the Taneja Group. Dell has put together popular SMB software with its own hardware, he said.

"These are good combinations, and I think through the hands of Dell, they will do very well," Taneja said.

The product, with as much as 144TB of usable disk space and a starting price between $10,000 and $15,000, would be best suited to organizations with a few hundred employees, which often have very limited IT skills on hand, Taneja said. As an all-in-one system, it should be most attractive to customers who haven't yet invested in another vendor's backup products, he added.

In addition to the choice of CommVault or Symantec Backup Exec software, users will have a choice of versions for each. A CommVault-based DL2000 can be configured with or without data de-duplication, which makes more efficient use of storage capacity by taking advantage of duplication in the incoming data. There are four editions of the Symantec software, which can provide continuous data protection for Exchange, SQL, and file servers, as well as protection for VMware and Microsoft virtual machines.

Each DL2000 also requires client software licenses for the systems that will back up their data to it. The CommVault Standard Edition comes with five licenses, and the Symantec Starter Edition comes with two.

To get started, IT staff will not have to set up LUNs (Logical Unit Numbers) or do RAID configurations, and it's possible to later add disks to the array on the fly. The systems will come with 3TB of disk capacity to start. Each version of software provides a management console that makes jobs such as setting up policies and backup tasks easy, according to Dell.

Archiving software is available for automatically moving data from disk to tape and setting up tiered storage. This feature, and the optional tape libraries, ought to be useful to medium-size enterprises with regulatory requirements for long-term storage, Taneja said.

AMS's decision to spin off its chip-manufacturing business into a separate company just may save the floundering firm, analysts said.

AMD's decision to split into a chip-designing company and a chipmaking company should help return the company to profitability sooner, said several analysts who watch the chipmaking industry. AMD would remain as the chip-designing company, while the new company, tentatively called The Foundry Co., would focus on chipmaking under the plan, unveiled Tuesday.

[ See this related story on AMD splitting up into two companies. ]

"I think this is a good move for AMD, especially in the current financial environment," said Dean McCarron, principal analyst for Mercury Research. "It frees AMD of the heavy-debt burden of owning a fab and lets it focus on its primary microprocessor and graphics businesses."

The move should allow AMD to return to profitability much faster than if it kept its foundry business, and it allows AMD to stop worrying about underuse of factories during economic downturns, McCarron added.

But the move is not without risks, added Jack Gold, founder and principal analyst at J.Gold Associates. The Foundry Co., despite a huge investment from the Advanced Technology Investment Company (ATIC), faces a competitive marketplace, Gold said. ATIC is a company set up by the government of Abu Dhabi to invest in tech companies worldwide. It will invest $1.4 billion directly in The Foundry Co. and pay another $700 million to AMD, giving it 55.6 percent of the new company. AMD will own the rest of the company.

The Foundry Co. will assume about $1.2 billion of AMD's debt. AMD's revenue from continuing operations for the second quarter of 2008 rose 3 percent from a year earlier, to $1.35 billion. However, the company posted a net loss of $1.19 billion for the quarter.

"AMD gets a sorely needed cash infusion," Gold said. "But the new company will need to get business in the door from other players, not just AMD, if it is to be successful longer-term. While semi-outsourcing to foundries is a major trend in the industry, it is also highly competitive and not always a good margin business."

Most analysts suggested the deal will put some added pressure on AMD competitor Intel.

"This is good for Intel in that it will offer renewed competition for Intel, which will in turn require it to maintain market leadership in designs and especially in its fabs," Gold said. "Intel has not always been the low-cost producer, and this deal will likely force it to see how it can cost-reduce and 'lean out' its manufacturing, which it has been doing already anyway."

Gold sees little potential for a negative impact to Intel, he said. "In fact, anything that makes AMD more competitive forces Intel to increase its own innovation, which is good for everyone -- Intel, system vendors, and consumers," he added.

The deal should put pressure on Intel to improve its technology, added Roger Kay, president of Endpoint Technologies Associates.

"The deal makes AMD much more viable over the long run, which directly challenges Intel, and that it will chase Intel into process manufacturing down at least to 22nm (nanometers) both as a chip company and as a foundry," Kay said. "The deal also puts other Intel competitors, like Nvidia, in a better position by supplying a great alternative foundry."

Intel doesn't yet know enough about the deal to understand how it will work, said Intel spokesman Chuck Mulloy. Mulloy called on AMD to allow the details of a confidential cross-licensing agreement between the two companies to be released to the public so that investors can understand the relationship.

Intel will stick with its approach of both making and designing chips, Mulloy said. An integrated design and manufacturing process can reduce time to market and result in better quality and lower costs, he said.

"We have long believed there are significant advantages to being an integrated device manufacturer," he added.

The deal could have a larger impact on other chip foundries, such as Taiwan Semiconductor Manufacturing Co. (TSMC), which supplies chips for AMD's ATI graphics processor arm, said Kay and Gold.

ATI's business with TSMC will continue for now, Gold said. "But eventually I would expect AMD to consolidate its designs around one fab process and one manufacturer," he added. "With the large investment they are making in CPU designs with the new foundry company, I expect that they will ultimately move their GPU designs over as well, cutting business with TSMC substantially."

The financial world seems to be falling apart, but Advanced Micro Devices has just landed several billion dollars from two Abu Dhabi investment groups to fund "Asset Smart," a plan to spin off its chip manufacturing operations into a new company, tentatively called "The Foundry Co."

The company still has to convince regulators and shareholders, but the move is already getting an enthusiastic response from customers and employees in Europe, according to company representatives.

[ See this related story on AMD splitting up into two companies. ]

"It's one of the most difficult weeks in the history of capitalism," said Emilio Ghilardi, senior vice president and general manager of AMD's sales and marketing activities in Europe, Middle East, and Africa (EMEA).

"From AMD's point of view, this is a great time because within this financial turmoil, we can say to our clients and our employees that AMD is a much stronger company," he said.

By spinning off its semiconductor fabrication plants, or fabs, into a new company and selling a majority share of that to one Abu Dhabi investment fund, and selling a smaller stake in its continuing business of processor design and marketing to another, AMD hopes to have greater financial freedom to invest in new product development.

"Talking to our key customers in Europe, I only got enthusiastic responses," Ghilardi said.

Ghilardi will stay with what he referred to as "AMD 2.0," the part of the company that will design and market the processors.

That company will retain some manufacturing activities related to testing and packaging, the final stages before the processors are shipped to customers.

"We can say to our customers 'Business as usual'," said Ghilardi, adding that there will be no change in the way the company manages customers orders or priority shipments.

On the other side of the coming dividing line will be Jens Drews, currently AMD's director of government relations for EMEA and a longtime resident of the German city of Dresden, the site of AMD's European semiconductor fabs.

"I've thrown my hat into the Foundry Co. side," said Drews.

Based in Europe, Drews won't be involved in one of the most pressing tasks the new company faces in government relations: negotiating the transfer of subsidies that the state of New York agreed to give AMD in return for building a new fab there.

The Foundry Co. plans to upgrade one of its two existing fabs in Dresden. Drews said he sees no need to seek fresh approval for state subsidies already granted for that operation.

Workers at that plant are taking the changes in their stride.

"In the cafeteria today, I sensed a growing excitement," he said. "It gives us a new opportunity," he said.

An ambitious project is under way in Wales to build one of the most advanced and secure datacenters in Europe.

Next Generation Data (NGD) is investing £200 million ($346 million) to modify a 750,000-square-foot (69,677 square meters) factory built by LG more than a decade ago near Newport, Wales, intended for fabricating microchips but abandoned after the Asian financial crisis.

The Welsh government has tried for years to find other uses for the hulking structure whose second floor alone can hold two 747 jumbo jets. Now, it's in the midst of a massive retrofit to house thousands of server racks that NGD officials say could appeal to companies such as IBM and Microsoft, which are growing their datacenter services.

Although it was built for microchips, NGD officials say the facility is well suited for a datacenter, ranging from its close proximity to local carrier exchanges, a nearby power substation and to the London area, said Simon Taylor, the company's chairman.

The datacenter is being designed to be "carrier neutral" and will have fiber Internet connections provided by carriers BT, NTL, and Cable and Wireless. NGD is also in negotiations to get a trans-Atlantic connection that would bypass London, which would ensure connectivity in the event of a terrorist attack or natural disaster such as flooding in the city.

"We actually think that having non-London transit will be a very saleable product for us," Taylor said.

NGD's facility will draw power from a local substation that connects directly to Britain's national grid, also called the "supergrid." The substation can deliver up to 180 megavolt amperes (MVA) in 45-MVA blocks. The amount of power is massive: 180 MVA is enough for a city, Taylor said.

It's tough to get that much power at datacenter sites within the city limits of London. "Power is a very, very scarce commodity," Taylor said.

The former LG facility had several traits that lend themselves well to becoming a datacenter. It has double-skin walls and was designed to be shudder-proof in order to make microchips.

NGD is also fortifying the facility to make it highly secure. A 12-foot (3.66-meters), military-grade fence extending into the ground will ring the facility with an infrared system beyond that to detect intruders. It will have thousands of closed-circuit television cameras, air-lock gates, retina-scan biometric entry systems, and bulletproof and bomb-proof reinforcements in places.

"We've really gone to town on this," Taylor said.

The walls will resist fire for up to two hours. In the event of a power failure, 18,000 liters of diesel fuel is stored on site, which would keep the center running for some 36 hours before more fuel would be needed. It has been classified as a Tier 3 facility by the Uptime Institute, which rates datacenters on infrastructure reliability and performance.

NGD has opted to only hire qualified ex-military personnel for its on-site guard force, which will be at the facility 24 hours a day.

Perhaps the biggest obstacle for NGD is the site's location in Newport, about an hour-and-45-minute train ride from London. Some server-hugging IT administrators don't like their equipment being very far away from their offices. But Taylor said those concerns are less relevant given that datacenters are typically managed from afar anyway.

NGD is hoping to sign up companies for long-term contracts. The facility has a total of 18 halls on three floors. The building will have a special section for small to medium-size enterprises and one for companies that are a bit larger. Those companies could either bring their own equipment or use hosting services offered by NGD.

Hosting space starts at 2,500 square feet. For large enterprises, the facility has 10,000- and 20,000-square-foot halls. NGD said it can provide 1.5 kilowatts per square meter or 4 kilowatts per rack up to a maximum of 8 kilowatts per rack.

NGD is in a race against others to get its facility up and running, as there are two to three other large datacenters planned in the U.K. and more throughout Europe, Taylor said. NGD's first clients should move in around March 2009 on the facility's ground floor. So far, NGD has commitments from clients for about 10 percent of the datacenter's 400,000 square feet of usable rack space and aims to fill the whole center in three years.

Since other data U.K. datacenter projects are less further along in their development and companies are desperate for more space, "I think we got our timing right on this. It's just about getting to the point where they're using the broom cupboard," Taylor said.

Researchers have published a cryptographic algorithm and source code that could be used to duplicate smart cards used by several major transit systems, including Boston's Charlie Card and the London Oyster card.

Scientists from the Dutch Radboud University Nijmegen presented their findings during the Esorics security conference on Monday in Malaga, Spain. They also published an article with cryptographic details.

[ Learn how to secure your systems with Roger Grimes' Security Adviser blog and newsletter, both from InfoWorld. ]

Their presentations show how to circumvent the security mechanism of NXP Semiconductor's Mifare Classic RFID cards, which are widely used to provide access control to buildings and public transportation systems.

Researchers had previously reported that they had broken cryptography used in Mifare Classic, but NXP sued to prevent the information from being made public. The company asked for a lengthy delay to give customers enough time to secure their systems. However, a Dutch court ruled that free speech protected the researchers and that they shouldn't fall victim to mistakes made by the supplier.

Some details on the Mifare encryption were already public, but the publication of the Dutch paper marks the first time this information has been disclosed in a complete form, making it easy for a determined attacker to clone one of the cards.

In the paper, the researchers say they discovered the workings of the chip by analyzing communication between the chip and the reader. An RFID-compatible device, the Ghost, was made to function independently from a computer. Ultimately they were able to obtain the cryptographic protocol. One issue is that the reader has to communicate in a predictable way, which opened the path to data analysis.

When the mechanism was cracked, the scientists were able to crack keys in less than a second using a industry standard computer with only 8MB of memory. Given the state of technology in 1996, when the Mifare Classic was introduced, even then such a crack would take only minutes.

The Radboud University paper shows that researchers looked at another chip, the Hitag2, in order to crack the Mifare Classic. This chip was later introduced, but cracked some years ago. Since the information on this hack is freely available, this helped the researchers.

Security experts had expected source code to surface soon after complete details of the Mifare Classic hack were published. However, last week a Russian Web site featured source code and documents. According to Professor Bart Jacobs, one of the paper's authors, the code lacked the authentication mechanism required to clone a Mifare Classic card.

But on Monday, another paper released by German researcher Henryk Plotz does feature functioning source code. Jacobs said that this implementation could be used to build a working card.

The German group working on the Mifare Classic chip cracked its encryption by removing a Mifare chip from a card and then cutting layers off. By photographing each layer under a microscope and analyzing all the connections they discovered the workings of the chip and derived the algorithm.

The Netherlands is introducing a new nationwide transportation card, OV-chipkaart, in a multi-billion-dollar project. Despite criticism, the Mifare Classic chip won't be replaced before the new technology is introduced in 2009.

Webwereld Netherlands is an InfoWorld affiliate.

Pramati Technologies will release new software widgets for the enterprise by January that will allow users to collaborate from within the business applications that they are using.

Collaborations tools installed by enterprises are currently in silos, and separate from the main business applications, said Jay Raghavendra Pullur, founder and CEO of Pramati. Users working on an ERP, order management, or CRM application are less likely to use a separate collaboration tool, using instead informal channels such as e-mail and even the telephone for collaboration, he added.

Pramati hopes to get around this problem by making its social and collaboration widgets "pluggable" and available from within business applications. The widgets from Pramati will support any business applications with a browser interface, and will work with collaboration tools like Microsoft's SharePoint already installed in an enterprise, Pullur said.

The widgets to be released for enterprise applications include a widget for project management, and another for discussions among users running different applications, according to Pullur. The widgets will figure in a portal from where users can download and paste code into applications, he added.

Pramati's first product was a J2EE application server that has been deployed by a number of customers worldwide.

Last month, Pramati's SocialTwist business unit launched its Tell-a-Friend Web widget, which lets readers recommend and share content directly with their friends through a variety of methods including e-mail, instant messaging, social networks, Twitter, or their blogs. Tell-a-Friend gives word-of-mouth marketing capability to Web sites, Pullur said.

The widget is offered free at www.socialtwist.com for users to copy and paste the code into their Web sites or blogs. It has already close to 2,000 installations, Pullur said.

The immediate objective of Pramati is to boost adoption of the Tell-a-Friend widget. Down the line it will look at advertising as a revenue stream from the widget.

A paid corporate edition of the widget, which does not have advertisements, is already being offered, and gives corporate users more control over the widget, and better context integration, Pullur said. Pramati already has seven corporate customers, he added.

The Tell-a-Friend widget is based on SaaS (software as a service) technology, which helps Pramati add new features to the widget quickly. There have already been requests for support for more blogging platforms, multiple languages, and other features, Pullur said. Pramati plans to release weekly upgrades that add new features to the widget.

The company also plans to offer more widgets for Web applications, including one that allows owners of Web sites to identify content on their site that they would not mind reused and blogged. Currently bloggers are never certain if they can use content from a Web site, without attracting copyright restrictions, Pullur said.

Advanced Micro Devices plans to split into two companies, one to design chips and one to make them, while two investment funds owned by the government of Abu Dhabi will contribute new capital, it said Tuesday. AMD hopes the move will give it the resources it needs to compete effectively with Intel, which dominates the microprocessor industry.

The Advanced Technology Investment Company (ATIC), created by the government of Abu Dhabi, will buy a substantial stake in the chip-making operation, tentatively called The Foundry Co. and contribute additional funds over the next five years to build a new chip fabrication plant, or fab, in New York state and to upgrade one of two AMD fabs near Dresden in Germany. The Foundry Co. will remain headquartered in the U.S., and will also make chips for other companies.

[ Stay ahead of advances in hardware technology with InfoWorld's Ahead of the Curve blog and newsletter. ]

ATIC will invest $1.4 billion directly in The Foundry Co. and pay another $700 million to AMD, giving it 55.6 percent of the new company and half the seats on the board. AMD will contribute intellectual property rights and its Dresden fabs to the company. AMD will own the remaining 44.4 percent and control the other half of the seats. The Foundry Co. will also assume around $1.2 billion of AMD's debt.

Over the next five years, ATIC will invest between $3.6 billion and $6 billion to fund redevelopment of one of the Dresden fabs and build a new one in New York.

At the same time, Mubadala Development Co. will pay $314 million to increase its stake in AMD to 19.3 percent. It bought an 8.1 percent stake in the company for $622 million last November.

The Foundry Co. will be led by Doug Grose, currently AMD's senior vice president of manufacturing operations, who will resign that post to become CEO of the new company. Hector Ruiz will resign as chairman of AMD to become chairman of the chip business. Ruiz recently quit as CEO of AMD, handing over the reins to Dirk Meyer who took over as president and CEO in July.

The companies expect to close the deal at the beginning of next year, if they get approval from AMD shareholders and regulators including the Committee on Foreign Investment in the United States.

AMD has been talking for some months about adopting an "asset light" or "asset smart" strategy in which it would divest some its debt-heavy manufacturing assets to focus on chip design.

While Intel and AMD have clung to their manufacturing facilities, "fabless" semiconductor companies such as ARM, a designer of microprocessors for mobile and embedded devices, and contract manufacturing companies such as Taiwan Semiconductor Manufacturing Co. (TSMC) are an established feature of the chip manufacturing industry.

AMD will announce its third-quarter earnings on Oct. 16, after the stock market closes. It has made a loss in the last seven quarters.

Even in the nuts-and-bolts world of tech execs, choosing a right-hand man (or woman, of course) can be as politically challenging as anything going on in the halls of Washington, D.C.

The right right-hand person can help your high-tech projects soar to new heights today and take care of your legacy when you leave tomorrow. The wrong one, though, can blur your technology vision to the brass, rankle the rank-and-file, and, worst of all, under-deliver on sweeping technology projects and initiatives -- in other words, make your life a living hell.

[ Read about the dismal state of IT workers, plus how the financial crisis affects IT jobs. | Seeking IT career advice? Rely on Bob Lewis' Advice Line blog at InfoWorld. ]

So how do you pick a running mate, a prot?g?, a No. 2, someone you can depend on? Perhaps the most important quality this person should possess is an absolute ability to execute, a hands-on pro who grasps the nuances of both technology and diplomacy, and gets stuff done.

The role of the No. 2 "The No. 2 keeps the trains running, while the CIO is concerned about where the trains are going," says Martha Heller, managing director and recruiter of technology leaders at executive search firm ZRG. "This person needs to fit in culturally with the organization and have the respect and relationships across the board to be able to fill in when the CIO is traveling -- essentially representing the CIO."

But choosing a No. 2 is fraught with risks. Too often, tech execs value a candidate's industry experience or a particular technology skill set over broader technology expertise and soft skills. A No. 2, for instance, may need to be able to succinctly summarize technical information to the board of directors without using mind-numbing acronyms. Tech execs also sometimes fail to spot obvious cultural misfits, such as a candidate who likes to run roughshod in a family-owned company.

Another role of the No. 2 is to take over if the tech exec moves up the corporate ladder, leaves for another company, retires, or is hit by a bus. In these cases, a good No. 2 is ideally suited to move into the No. 1 spot. "They have a view into each part of the organization so that their perspective is broad enough if there comes a need for a successor," Heller says.

To be fair, most CTOs and CIOs don't give such succession planning much thought. A few years ago, staffing firm Robert Half Technology surveyed some 1,400 CIOs and found that only one in five did succession planning.

Tech skills versus soft skills Chuck Kramer, senior vice president and CTO of Social and Scientific Systems, doesn't have a successor yet. But he does have eight right-hand folks, each overseeing a technical area: security, enterprise systems, software development, bioinformatics, and three operation sites. Two out of the eight are under consideration as his successor, although they still have a ways to go. "What they need, and what is often lacking, is assistance in strategic planning," Kramer says.

It's an endemic problem when looking for a No. 2, agrees Heller. And that's why the vast majority of times, she says, a tech exec will look outside of the organization for a right-hand person because internal candidates lack strategic thinking. "The [inside] bench often has narrowly focused technology skills and not enough executive presence," she says. "Also, by going outside, there's an opportunity to bring another perspective to your IT organization. It's a very valuable moment."

On the technology front, Kramer's eight deputies are all pretty technically savvy -- and not just in their area of expertise. When choosing them, he made sure that they could explain how their projects would fit into the overall business. "They must see the big picture, rather than only their fiefdom," Kramer says.

A bad No. 2 who is in love with a particular technology may not be able to adapt to the changing needs of an IT organization or even make fair decisions about the technology stack that the tech exec has spent years building. Love for a particular technology vendor can be just as wrongheaded. On the other hand, a good No. 2 candidate should be able to show leadership in vendor reviews, SLA enforcement, and cost savings.

Mirror image or mirror for your blind spots? Such well-rounded candidates are rare. Ten years ago, CTO Matt Kesner of law firm Fenwick and West cherry-picked networking guru Kevin Moore at Novell to help shore up the law firm's network. Kesner, a former lawyer and programmer, needed a technician to complement his visionary skills. (Indeed, Kesner recently created a new practice at Fenwick and West as a result of that vision.)

What Kesner really got was a No. 2 in the making. Fenwick and West needed someone with deep technology skills, but over time, Moore developed into a diplomat, manager, and leader, Kesner says. Moore inherited a downtrodden staff and inspired them by becoming a role model. "Given my background, it was hard for [IT workers] to see me as a future of themselves," Kesner says. "With Moore, they saw what they could be. He's a natural leader."

ZRG's Heller says tech execs should look to fill technology blind spots with their No. 2. For example, if the tech exec climbed the ladder through infrastructure and doesn't have a strong foundation in applications, then he may want to bring in someone specifically on apps.

But this need to fill in blind spots doesn't mean that the No. 2 should be vastly different from the tech exec in other areas. "You should look for somebody who is a mirror image of you in core values of leadership," says Heller. "For example, if you're about treating the business as a client, then you need someone to mirror that."

Today, Moore is IT director and fills in for Kesner from time to time, performing CTO duties and making critical technology decisions. If Moore doesn't leave for a CIO or CTO position somewhere else, he'll likely succeed Kesner. Moore's dilemma is that even though he is the obvious successor to Kesner, he and Kesner are roughly the same age, and Kesner doesn't have plans to leave Fenwick and West anytime soon.

The power of ambition Tech execs must choose whether they want a rising star or a conventional technologist to be their No. 2. Vice presidential candidate Sarah Palin's recent gaffe calling the ticket "Palin and McCain" rather than McCain-Palin may have been simple mistake -- or a Freudian slip. An ambitious No. 2 who dreams of being top dog one day isn't necessarily a bad thing, but a tech exec should make sure that her right-hand man is focused on present-day execution.

Because tech execs don't have term limits and can hold office for many years, an ambitious understudy may become frustrated by the lack of upward mobility and turn into a flight risk. To avoid this possibility, Kesner and Moore regularly discuss five-year plans and increased responsibilities, such as Moore handling CTO duties.

When evaluating a candidate for a No. 2 post, tech execs should ask where the candidates see themselves in five years. Are they seeking new skill sets to move up the corporate ladder? Or do they think they're already capable? This will give some indication about how the candidates view the runway at a company and their own career aspirations.

And it's not bad to develop No. 2s who end up as No. 1s elsewhere. "If you become a place that trains great leaders, that behooves you," Heller says. After all, it not only increases your network, but also earns you respect in your organization and enhances the image of IT.

Microsoft laid out on Monday its road map for SQL Server and a complement of add-ons it hopes will eventually redefine business intelligence and data warehousing.

The next version of the database server, code-named Kilimanjaro, is slated for release in the first half of 2010 with a focus on self-service and reporting capabilities for BI. Microsoft plans to have a "community technology preview" (CTP) available within the next 12 months.

The self-service features are wrapped up in a set of technologies code-named Gemini. Those technologies let users build BI applications that can access data across many sources, aggregate the data, build charts and reports, and share the resulting applications via SharePoint.

Microsoft also plans to integrate the unified communications capabilities of Office Communications Server to aid the sharing of BI results.

Microsoft said much of the Gemini technology will be tied to Excel, allowing users of that desktop program access to the self-service analytics.

Microsoft made the announcements at its annual BI Conference, which is going on in Seattle this week and is expected to draw 2,500 users and partners.

The company's acquisitions in the BI market and its stated intentions to expand BI capabilities on the back of its popular SQL Server have been shaking up the market in the past six months.

Giants like Business Objects/SAP, Cognos/ IBM, and Hyperion/ Oracle are among the BI heavyweights with an eye on Microsoft.

A report by Gartner earlier this year said Microsoft still "lags behind pure-play vendors in terms of metadata management, reporting, and dashboard and ad hoc query capabilities."

Microsoft plans to systematically address those deficiencies.

The immediate goal is to extend its BI tools and software so they are more accessible to users, especially those using Excel and SharePoint.

With Gemini, Microsoft hopes to bring BI to users without sacrificing IT control.

"One important thing about Gemini is managed self-service," says Fausto Ibarra, director of product management for SQL Server. "Managed means IT is in control of the process where today end-users use Excel without control of IT or without control on data."

With Gemini, IT will be able to see how data is being shared, will have control of security on the data, and will make data sources available to users.

Those sources could include ERP data, mainframe applications and independent software vendor programs.

Another key feature of Gemini is in-memory BI, which analyzes large amounts of data in memory in order to speed performance.

At the conference, the company also unveiled plans for a highly scalable database technology code-named Madison that would be available in an appliance. Madison integrates SQL Server with technology the company acquired when it bought DataAllegro earlier this year.

DataAllegro developed large-volume data warehousing appliances, and Microsoft hopes to scale Madison to handle hundreds of terabytes of data. At the conference, Microsoft showed a demo using 1 trillion rows of data.

The company also plans to use data quality technology acquired when it bought Zoomix in July to enhance the quality of available information. Microsoft would only say the technology will come in "future versions" of SQL Server.

CTPs of Madison will roll out in the next 12 months with the appliances available in the first half of 2010. Dell, HP, Unisys, Bull Systems, and EMC have signed on as hardware partners.

Microsoft laid out on Monday its road map for SQL Server and a complement of add-ons it hopes will eventually redefine business intelligence and data warehousing.

The next version of the database server, code-named Kilimanjaro, is slated for release in the first half of 2010 with a focus on self-service and reporting capabilities for BI. Microsoft plans to have a "community technology preview" (CTP) available within the next 12 months.

The self-service features are wrapped up in a set of technologies code-named Gemini. Those technologies let users build BI applications that can access data across many sources, aggregate the data, build charts and reports, and share the resulting applications via SharePoint.

Microsoft also plans to integrate the unified communications capabilities of Office Communications Server to aid the sharing of BI results.

Microsoft said much of the Gemini technology will be tied to Excel, allowing users of that desktop program access to the self-service analytics.

Microsoft made the announcements at its annual BI Conference, which is going on in Seattle this week and is expected to draw 2,500 users and partners.

The company's acquisitions in the BI market and its stated intentions to expand BI capabilities on the back of its popular SQL Server have been shaking up the market in the past six months.

Giants like Business Objects/SAP, Cognos/ IBM, and Hyperion/Oracle are among the BI heavyweights with an eye on Microsoft.

A report by Gartner earlier this year said Microsoft still "lags behind pure-play vendors in terms of metadata management, reporting, and dashboard and ad hoc query capabilities."

Microsoft plans to systematically address those deficiencies.

The immediate goal is to extend its BI tools and software so they are more accessible to users, especially those using Excel and SharePoint.

With Gemini, Microsoft hopes to bring BI to users without sacrificing IT control.

"One important thing about Gemini is managed self-service," says Fausto Ibarra, director of product management for SQL Server. "Managed means IT is in control of the process where today end-users use Excel without control of IT or without control on data."

With Gemini, IT will be able to see how data is being shared, will have control of security on the data, and will make data sources available to users.

Those sources could include ERP data, mainframe applications and independent software vendor programs.

Another key feature of Gemini is in-memory BI, which analyzes large amounts of data in memory in order to speed performance.

At the conference, the company also unveiled plans for a highly scalable database technology code-named Madison that would be available in an appliance. Madison integrates SQL Server with technology the company acquired when it bought DataAllegro earlier this year.

DataAllegro developed large-volume data warehousing appliances, and Microsoft hopes to scale Madison to handle hundreds of terabytes of data. At the conference, Microsoft showed a demo using 1 trillion rows of data.

The company also plans to use data quality technology acquired when it bought Zoomix in July to enhance the quality of available information. Microsoft would only say the technology will come in "future versions" of SQL Server.

CTPs of Madison will roll out in the next 12 months with the appliances available in the first half of 2010. Dell, HP, Unisys, Bull Systems, and EMC have signed on as hardware partners.

Mixing together a m?lange of services, software, and marketing, IBM's announcement this week of its Cloud Services Initiative is about putting an organizing construct around all of its cloud offerings, according to one IBM executive.

To that end, it does not appear that there is much new in the way of products or services in the initiative.

Mostly repackaging of IBM datacenter and Lotus technology Bluehouse, the centerpiece of the initiative, is a Web-based tool for collaboration. However, while the name may be new, Bluehouse actually incorporates a great deal of Big Blue's existing products, such as IBM's Lotus SameTime collaborative and social networking environment. Bluehouse builds on the services currently offered in SameTime for instant messaging and unified communications. "Bluehouse adds document sharing, contact sharing, and community building all in a SaaS (software-as-a-service) model," said Dave Mitchell, director of strategy for cloud services at IBM.

[ Confused about what cloud computing really means? Find out in InfoWorld's definitive guide to cloud computing. And for more on Bluehouse, see "IBM bundles up cloud computing initiatives." ]

Along with Bluehouse, IBM also announced SameTime Unyte, a Web conferencing offering. Unyte is part of Bluehouse as well.

In addition to Bluehouse, the four-part initiative adds to IBM's SaaS platform offerings, whereby IBM hosts the delivery infrastructure for software vendors. What's new is not technological but marketing: IBM has broadened the definition of a partner, expanding it to mean any software vendor that uses two out of three products IBM delivers: middleware, hardware, and managed hosting. Partners are also the beneficiary of joint marketing efforts.

Services to integrate cloud components into an organization's business environment are also available for IBM customers.

The fourth component will provide a datacenter environment based on IBM's Cloud Computing Centers around the world. This will give customers remote access to computing power in an on-demand environment.

Looking past proprietary clouds Although the Bluehouse effort appears to be something IBM has been doing for a considerable number of years through its datacenter services, Mitchell hinted at doing something more, resolving an issue that has recently been swirling around the use of cloud solutions. "We are working with our partners using SOA to develop open clouds as opposed to proprietary clouds as in the past."

This comes on the heels of recent statements by Richard Stallman, for example, founder of GNU and a well-respected industry watcher who was quoted in the British newspaper The Guardian as saying cloud computing is nothing more than a "marketing hype campaign" and a "trap which will lock users into proprietary systems."

The difference between what IBM is offering and others, says Mitchell, is that an open cloud environment will give users more interoperability and more connections to partners.