November 12th, 2009

PDC 2009: Tune in for our live blogging frenzy next week

Posted by Mary Jo Foley @ 6:31 am

Categories: .Net Framework, Azure, Corporate strategy, Development tools, Internet Explorer, Microsoft Big Brains, Multicore/distributed computing, Office 2010/Office 14, PDC 2009, Red Dog, Research, Silverlight (wpf/e), Utility/cloud computing, Virtualization, Visual Studio 10 ("Hawaii"), Windows 7, Windows Server 2008 R2 /("Windows 7 Server")

Tags: Microsoft Corp., Professional Developers Conference Keynote, Blogging, Internet, Mary Jo Foley

Microsoft’s Professional Developers Conference (PDC) 2009 kicks off the week of November 16. Like we did last year, a handful of us Microsoft watchers will be live blogging the keynotes as a group.

The PDC keynotes are slated for Tuesday November 17 from 8:30 a.m. to 10:30 a.m. PT and Wednesday November 18 from 8:30 a.m. to 10:30 a.m. We’ll be using CoverItLive to blog, so the more of you who chime in and comment along with us, the merrier. Your group-blogging hosts (besides me) will be Ed Bott, Kip Kniskern, Paul Thurrott, Rafael Rivera, Tom Warren and Long Zheng

Come back here next week and watch along with us as Chief Software Architect Ray Ozzie; Server and Tools President Bob Muglia; Kurt DelBene, Senior VP of Microsoft’s Office Business Productivity Group and more talk about what’s coming for developers in the next year. (I’ll post the CoverItLive viewer on my site during keynote viewing hours next week.)

There will be new info on Microsoft’s Azure cloud operating environment, .Net 4.0, Oslo, Office 2010, Silverlight, SQL Server and more. And more than a few of the “Big Brains” — Microsoft’s Technical Fellows — are on tap to present during the four-day confab.  I’ve already posted about some of what’s on tap (and not on tap) for PDC 2009 over the past few weeks. Expect lots more PDC news on my blog throughout the week next week.

Hope to see you (virtually) and/or live in Los Angeles next week!

October 19th, 2009

Testers to get Visual Studio 2010 Beta 2 this week; final by March 2010

Posted by Mary Jo Foley @ 6:44 am

Categories: .Net Framework, Azure, Corporate strategy, Development tools, Multicore/distributed computing, SharePoint Server, Silverlight (wpf/e), Utility/cloud computing, Visual Studio 10 ("Hawaii"), Windows 7, Windows Server 2008 R2 /("Windows 7 Server")

Tags: Microsoft Developer Network, Microsoft Corp., Beta, Microsoft Visual Studio, Microsoft Development Tools, Development Tools, Software Development, Software/Web Development, Mary Jo Foley

Microsoft is making what may be the last beta of Visual Studio 2010 and the accompanying .Net Framework 4 before they launch next March available to testers this week, company officials said.

MSDN testers will be able to download Beta 2 on October 19. Microsoft plans to open the beta to the public on October 21. The company is planning to launch the final version of its latest development suite on March 22, 2010, officials said. Microsoft’s goal is to deliver the actual bits by that date, not just to hold a launch.

Microsoft launched Beta 1 of Visual Studio 2010 and .Net 4 in May.This past summer, Microsoft officials told partners to expect the marketing/training/sales push for Visual Studio 2010 to begin in April 2010, so it sounds like the development is running on schedule.

Microsoft is positioning Visual Studio 2010 as its tool platform to support Windows 7, Windows Server 2008 R2, Azure, SQL Server, Office 2010 and SharePoint 2010. Support for SharePoint 2010 is new, as of Beta 2, officials confirmed. SharePoint is the “fastest growing platform, from a developer mindset,” for Microsoft at this point, said Dave Mendlen, Senior Director of Developer Marketing.

Visual Studio 2010 also includes new drag and drop bindings for Silverlight and Windows Presentation Foundation; interoperability with the ASP.Net model view controller (MVC), better multicore support and UML support.

Microsoft is touting .Net 4 as being 81 percent smaller than its predecessors, making it quicker and easier to download and install. Also unlike its predecessors, .Net 4 can be installed side-by-side with the previously released .Net 3.5. It adds support for the Microsoft Dynamic Language Runtime (DLR), giving programmers more language choices; and is more suited for parallel-programming, workflow-centric and service-oriented application development, according to the company.

“Beta 2 is not about dramatic changes to the features but is more about improvements to the performance and quality,” said Soma Somasegar, Senior Vice President of Microsoft’s Developer Division.

Microsoft officials also shared on October 19 more details about the planned packaging and pricing for Visual Studio 2010. Microsoft is cutting the number of SKUs of Visual Studio to four main ones, and is doing away with the database, architect and test versions. The four:

• Visual Studio 2010 Ultimate with MSDN. Includes all of the current Visual Studio Team System functionality. $11,924 for a new license ; $3,841 for a renewal
• Visual Studio 2010 Premium with MSDN. $5,469 new; $2,299 renewal
• Visual Studio 2010 Professional with MSDN. $1,199 new; $799 renewal
• Visual Studio 2010 Professional without MSDN. $799

MSDN subscribers will be getting unlimited access to Visual Studio Team Foundation Server 2010 (upon release), its team-collaboration server; a set (variable) number of compute hours per month for Windows Azure development; and up to 40 hours per year of e-learning classes per subscriber.

To attempt to get developers to move to MSDN Premium before Visual Studio 2010 launches, Microsoft has created the Ultimate Offer for VS developers. Anyone who is an active subscriber to MSDN Premium by the time Visual Studio 2010 launches next March will be transitioned automatically to the next higher level VS 2010 SKU with an MSDN subscription at launch.

October 5th, 2009

Microsoft's Midori: Who's on the all-star roster?

Posted by Mary Jo Foley @ 8:11 am

Categories: .Net Framework, Code names, Corporate strategy, Development tools, Multicore/distributed computing, Research, Virtualization, Windows client

Tags: Operating System, Blog, Microsoft Corp., Midori, Operating Systems, Software, Mary Jo Foley

It’s been a while since anything new about Microsoft’s Midori project has leaked. But thanks to a post on the “Codename Windows blog” plus a little poking around, I found an interesting list.

Microsoft officials have repeatedly refused to talk about Midori, other than to admit it is an incubation project (and with the disclaimer that it may never see the light of day). For a project that may never materialize, Midori seemingly has some heavyweight talent behind it.

First, a quick recap: Midori is all about building a new operating system that isn’t based on the current Windows kernel. Headed by Senior VIce President of Technical Strategy Eric Rudder, Midori is/was slated to be a distributed, concurrent operating system, according to various tips.

Rob Jellinghaus — a Principal Architect at Microsoft “working on an unannounced incubation project” — posted to his blog on September 11a “list of worthy programmers.” Jellinghaus doesn’t ever state that these folks are working on Midori, but he does note that he is part of a team that “working on a new operating system stack from boot loader all the way to applications. I can’t really say much more, except that what we’re doing is not entirely unrelated to the Singularity operating system.” Sure sounds like Midori to me….

Early leaks about Midori indicated Midori had roots in the Singularity microkernel operating system developed by Microsoft Research. Low and behold, a number of the programmers on Jellinghaus’ “worthy” list have worked on Singularity, as well as on other distributed operating systems, compilers and other related components. (Jellinghaus himself was “one of the first outside contributers to the Google Web Toolkit. He also worked on the Xanadu hypertext system.)

On Jellinghaus’ list:

•Daniel Lehenbauer: Describes his role on the unnamed Microsoft incubation project — which he calls the “most exciting and revolutionary work to happen in the industry since (Xerox) PARC” — as involving “the exploration of a radically different approach to the UI/Graphics platform which guarantees security, responsiveness, and leverages modern GPUs and manycore.” Software Design Engineer Lehenbauer says the incubation team of which he is a part is “revisiting every layer of the stack from device drivers, through rendering engines, up to application frameworks and programming/computation models.”

• Pavel Curtis: Software Architect, who, according to his profile on Wikipedia, “is best known for having founded and managed LambdaMOO, one of the best-known online communities of the 1990s. He created LambdaMOO during his 13-1/2 years as a member of the research staff at Xerox PARC, from 1983 to 1996, where he worked in the areas of programming language design and implementation, programming environments, and online collaboration systems.”

• Jonathan Shapiro: One of the chief developers of the BitC language and Coyotos operating system, joined the Midori team this past spring, he acknowledged in a blog post.

• Ravi Pandya:  An “Architect, Technical Strategy Incubation,” according to his blog profile. From a 2007 blog post: “I moved from Windows Security to an incubation group which is, as Chris Brumme so eloquently puts it, ‘exploring evolution and revolution in operating systems.’ I’m having a lot of fun working with a variety of interesting systems technologies, including security, distributed systems, many-core, virtualization, managed systems code, dynamic resource scheduling, asynchronous & adaptive user interfaces, etc.”

• Dean Tribble: A Principal Architect at Microsoft, Tribble led development of security and compliance features for Microsoft Exchange, and “now is incubating new operating systems technologies.”

• Chris Brumme: A Microsoft distinguished engineer who was an architect on the Common Language Runtime (CLR) team. More recently, Brumme “has been one of the architects on an unannounced systems project.”

• Bjarne Steensgard: Since 2007, has been “part of an incubation team at Microsoft that is an outgrowth of efforts started at Microsoft Research.” At Microsoft Research, he worked on the Marmot and Bartok compilers and runtime systems. (Bartok was influential in the development of Singularity, on which Steensgard also worked “since its inception,” he said. Bartok also seems to figure into the Midori picture.) Before joining Microsoft, he worked on the Emerald distributed operating system.

•David Tarditi: A former Microsoft researcher who worked on Singularity.

•Tanj Bennett: One of the 40-plus Softies running the revamped Microsoft ThinkWeek program. His area of specialization is “OS in the Future.” Bennett also seems to have a connection with a Microsoft Research project known as the “Microsoft Solver Foundation,” which is described as “a new framework and managed-code runtime for mathematical programming, modeling, and optimization.”

• Joe Duffy: The Lead Developer and Architect for Parallel Extensions to .NET. Author of the book Concurrent Programming on Windows

• Leif Kornstaedt: Worked for several years on the CLR as a developer and a senior development lead; now “work(s) in Technical Strategy Incubation.” His area of specialization, according to his Web page, is “design and implementation of a programmable middleware.” He contributed to Alice, a functional programming language, and Mozart, an implementation of the Oz language.

Midori has been in the works since 2006/2007, based on the bios of some of these individuals. But there’s no inkling of when it might emerge from incubation land. As I’ve reported before, Microsoft is working on a couple of related projects (codenamed “RedHawk” and “MinSafe”) that are supposedly pre-cursors to Midori and which could work their way, at least in part, into Windows 8.

September 28th, 2009

Helios: Another Microsoft operating system project to watch

Posted by Mary Jo Foley @ 6:58 am

Categories: Code names, Corporate strategy, Multicore/distributed computing, Research, Windows client, Windows server

Tags: Helios D.d., Operating System, Microsoft Corp., Singularity, Operating Systems, Software, Mary Jo Foley

Microsoft’s researchers are working on yet another operating-system research project which can trace its roots to the company’s Singularity project. This new operating system, known as Helios, is a heterogeneous multiprocessing platform built around satellite kernels.

(The folks over at the Ma-Config.com blog sent me a pointer to Helios after I wrote last week about another Microsoft Research operating system project, codenamed “Barrelfish.” Without the link they provided, I wouldn’t have found information about Helios, as it isn’t listed on the active projects page for Microsoft Research. Microsoft researchers have written a 14-page paper on Helios, however, which is slated for publication in October.)

Singularity, in case you need a quick refresher, is a microkernel operating system and set of related tools and libraries that is developed completely in managed code. Singularity is not based on Windows; it was written from scratch as a proof-of-concept. Microsoft’s Midori incubation project is another effort which can trace its lineage to Singularity.

What, exactly, is Helios? From the soon-to-be-published ACM paper about it:

“Helios is an operating system designed to simplify the task of writing, deploying, and tuning applications for heterogeneous platforms. Helios introduces satellite kernels, which export a single, uniform set of OS abstractions across CPUs of disparate architectures and performance characteristics. Access to I/O services such as file systems are made transparent via remote message passing, which extends a standard microkernel message-passing abstraction to a satellite kernel infrastructure. Helios retargets applications to available ISAs by compiling from an intermediate language.”

According to the paper, the team built Helios by modifying the Singularity research development kit (RDK) to support satellite kernels, remote message passing and affinity. They implemented satellite-kernel support on two different hardware platforms: an Intel XScale programmable PCI Express I/O card and cache-coherent NUMA architectures. Helios “treats programmable devices as part of a ‘distributed system in the small,’” according to Microsoft’s description, and “is inspired by distributed operating systems such as LOCUS, Emerald and Quicksilver.”

The Helios researchers describe Helios and Barrelfish, another Microsoft Research OS project, as complementary. From the paper:

“Barrelfish focuses on gaining a fine-grained understanding of application requirements when running applications, while the focus of Helios is to export a single-kernel image across heterogenous coprocessors to make it easy for applications to take advantage of new hardware
platforms.”

While there has been lots of talk about what Microsoft is planning to deliver as the successor to Windows, it’s worth remembering that Singularity, Barrelfish, Helios and Midori are all in early stages — and might not ever be commercialized. While Microsoft officials don’t mind talking about the office of 2019, they don’t want to share anything at all on the version of Windows expected in 2011/2012, let alone anything beyond that. So it’s tough to say how/if any of these future OS projects will influence the next big OS thing at Microsoft. Still, they’re all definitely worth watching….

September 25th, 2009

Microsoft and European researchers deliver a snapshot of multikernel 'Barrelfish' OS

Posted by Mary Jo Foley @ 8:14 am

Categories: Azure, Code names, Corporate strategy, Google, Multicore/distributed computing, Research, Utility/cloud computing, Windows CE

Tags: Operating System, Microsoft Corp., Barrelfish, Operating Systems, Software, Mary Jo Foley

Microsoft and European researchers have released a “snapshot” of Barrelfish, which is a “multikernel” distributed operating system project in which Microsoft’s Cambridge (UK) Research Lab has a big part.

Barrelfish isn’t a new project; Microsoft Research and ETH Zurich, a university focused on science and technology, have been working on it since at least 2007. But on September 14, the team released an early version of the operating system for those interested in testing it.

(The licensing/copyright terms for Barrelfish are here, along with a note that the Os includes “some third-party libraries, which are covered by various BSD-like open source licenses.”) The team also noted that “a more complete (and usable) release” of Barrelfish will be coming “soon.”

According to the Microsoft Research Web site, Barrelfish is one of a number of projects in which Microsoft is involved in the multi/many-core space. Here’s the description of the project:

“We are exploring how to structure an OS for future multi- and many-core systems. The motivation is two closely related hardware trends: first, the rapidly growing number of cores, which leads to scalability challenges, and second, the increasing diversity in computer hardware, requiring the OS to manage and exploit heterogeneous hardware resources.”

Microsoft and the ETH Zurich researchers have made available several  white papers outlining the goals and design of Barrelfish. None of those papers mention “Midori,” another distributed operating-system project that is in incubation at Microsoft, nor Singularity, another Microsoft Research operating-system project that inspired Midori.

But there still are a few interesting tidbits in the Barrelfish papers. The Barrelfish project may have implications for Microsoft’s Dryad and Google’s MapReduce, two programming initiatives that are key to the two companies’ datacenter/cloud efforts. (Here’s what I’ve written on Microsoft’s Dryad concurrent-computing work, if you want more background.)

From Microsoft’s latest Barrelfish paper:

“Structuring the OS as a distributed system more closely matches the structure of some increasingly popular programming models for datacenter applications, such as MapReduce and Dryad, where applications are written for aggregates of machines.”

And ARM processors are on the list of those slated to supported by Barrelfish. Microsoft’s Windows Embedded CE operating system runs on ARM but no variants of Windows do (at least for now).

Any OS experts out there see anything else of note in the Barrelfish papers? Any thoughts as to how it might fit (or not) with what we’ve heard to date about Microsoft Midori?

September 22nd, 2009

Microsoft buys desktop parallel-computing software maker ISC

Posted by Mary Jo Foley @ 5:44 am

Categories: .Net Framework, Corporate strategy, Development tools, Linux, Multicore/distributed computing, Windows client, Windows server

Tags: Software, Desktop, High-performance, Microsoft Corp., ISC, Leadership, Management, Mary Jo Foley

Microsoft has purchased the assets of Interactive Supercomputing (ISC), a desktop parallel-computing vendor, according to a blog post on the Windows Server Division Weblog.

The move fits in with Microsoft’s continued efforts to build up its high-performance server capabilities so as to better take on Linux in that market. It also fits with Microsoft’s larger Server and Tools and research projects to make parallel programming easier.

Microsoft has a wide variety of ongoing projects in the parallel/distributed computing space, from its Axum concurrent-computing language, to the Parallel FX extenions to the .Net Framework. Earlier this year, Microsoft formed a new eXtreme Computing Group, headed by supercomputing expert Dan Reed, that is focused on exascale computing.

According to the Windows Server post, dated September 21, from Kyril Faenov General Manager, High Performance & Parallel Computing Technologies:

“ISC’s products and technology enable faster prototyping, iteration, and deployment of large-scale parallel solutions, which is well aligned with our vision of making high performance computing and parallel computing easier, both on the desktop and in the cluster.”

Faenov said in the post that Microsoft recently began planning to integrate ISC technologies into future versions of unnamed Microsoft products and that Microsoft will provide more details “over the coming months.” In the short term, Microsoft will be providing support for ISC’s current Star_P customers, he said.

The CEO of ISC, Bill Blake, will be bringing the ISC team to work at Microsoft’s New England Research & Development Center in Cambridge, MA.

PEHub reported yesterday that Microsoft had purchased ISC, but Microsoft declined to confirm that report.

PEHub noted that ISC’s P-Star is “technical computing software that enables users to code computing problems on their desktops using familiar mathematical software such as MATLAB and Python, and run them instantly and interactively on parallel high-performance computers (HPCs).”

September 8th, 2009

Microsoft Doloto: Making Ajax page downloads faster

Posted by Mary Jo Foley @ 7:21 am

Categories: .Net Framework, Code names, Corporate strategy, Development tools, Multicore/distributed computing, Research, Web 2.0

Tags: Microsoft Corp., Tool, AJAX, WANs, Productivity, Networking, Internet, Software/Web Development, Web Development, Web 2.0

Microsoft is making available for download a tool developed by Microsoft Research that is designed to make Web pages more responsive by decreasing the initial download size of Ajax applications.

That tool, known as Doloto, analyzes Ajax app workloads and performs code splitting of large Web apps. That way, apps can transfer initially only the part of the client-size JavaScript code needed for application initialization. The rest of the function code is transferred later, according to Senior Vice President of Microsoft’s Developer Division Soma Somasegar, who blogged about Doloto on September 4.

Microsoft is moving the Doloto project to its Dev Labs incubator, home of a number of other developer-focused prototypes, including the Axum parallel-processing language and CHESS concurrency testing tool.

According to the Dev Labs site:

“Doloto reduces the size of initial application code download by hundreds of kilobytes or as much as 50% of the original download size. The time to download and begin interacting with large applications is reduced by 20-40% or dozens of seconds…depending on the application and wide-area network conditions.”

Doloto is short for “Download Time Optimizer” and is also the Russian word for chisel, according ot the Microsoft Research site.

August 4th, 2009

Microsoft PDC 2009: What's on tap for developers this November?

Posted by Mary Jo Foley @ 5:41 am

Categories: Azure, CRM Live/CRM Online, Channel, Corporate strategy, Database, Development tools, Live Mesh, Microsoft Big Brains, Multicore/distributed computing, Office 2010/Office 14, PDC 2009, Red Dog, SQL Server, SharePoint Server, Silverlight (wpf/e), Utility/cloud computing, Visual Studio 10 ("Hawaii"), WinHEC, Windows 7, Windows Mobile, Windows Server 2008 R2 /("Windows 7 Server"), Windows client, Windows server

Tags: Developer, Microsoft Corp., WinHEC, Microsoft Windows, Operating Systems, Mobile Operating Systems, Mobile Applications, Handhelds, Software, Hardware

Microsoft opened up registration for its Professional Developers Conference (PDC) 2009 on August 4. This year’s developer-fest is in Los Angeles again, as it was last year, and will run from November 16 to 19.

As it will be too early for Microsoft to start detailing publicly what’s coming in Windows 8, what will the Softies be highlighting at this year’s conference?

Keynotes are on tap from the elusive Chief Software Architect Ray Ozzie and Server and Tools President Bob Muglia. (Other keynotes will be added to the agenda later this year, company officials said.) Azure, Microsoft’s cloud platform, is slated to go from beta to final around the time of the PDC. And Microsoft also plans to offer a number of sessions on its hosted-development strategy at the show, as well. (Think xRM, SharePoint Services, and other “utility computing” building blocks.) Sessions dedicated to SQL Azure, Microsoft’s hosted SQL Server service, also are on the agenda.

Even though Microsoft will have launched Windows 7 and Windows Server 2008 less than a month before the PDC, on October 22, there will still be lots of sessions around developing for those platforms. Windows expert and Technical Fellow Marc Russinovich is on tap to talk about the kernel changes Microsoft made in Windows 7/Windows Server 2008 R2.

Russinovich isn’t the only “Big Brain” slated to speak at the confab. Other Microsoft Technical Fellows, including parallel-computing expert Burton Smith and database guru Dave Campbell are on the line-up, as well.

There is going to be a lot of content on .Net Framework 4 and Visual Studio 2010 (both of which are expected to launch around spring 2010, last I heard). Microsoft also is promising a “sneak peek” at some of the so-far unspecified new features in the next version of Silverlight (which I’d expect to be called Silverlight 4) during the PDC.

There are a few sessions dedicated to developing for SharePoint 2010 on the docket. There’s a session on “Office 2010 as a RAD (Rapid Application Development) platform.” I’m assuming this is Microsoft’s updated Office Business Applications (OBA) platform/strategy.

I asked Tim O’Brien, Director of Microsoft’s Platform Strategy Group, whether there would be much, if any, PDC content dedicated to developing for Windows Mobile. The first Windows Mobile 6.5 phones are set to launch in October and developers are champing at the bit for information on Windows Mobile 7.

“We are marching toward getting 6.5 out. That’s our focus right now,” is all O’Brien would say. In other words, guess we’ll have to wait and see if WinMo makes it onto the PDC agenda in any meaningful way.

I also asked O’Brien whether Microsoft was still planning to hold a Windows Hardware Engineering Conference (WinHEC) this year. Last year, Microsoft cancelled WinHEC.

“There’s no reason to say we won’t,” said O’Brien. But he also acknowledged he didn’t know for sure whether or when Microsoft would hold a WinHEC in 2009. Guess that’s another wait-and-see, too.

Microsoft is planning to Webcast the keynotes and many of the PDC sessions for those who can’t make the show in person. Me? I’m planning on making the cross-country trek myself to my favorite Microsoft show of the year.

July 28th, 2009

Microsoft releases second, 'experimental' version of .Net 4 Beta 1

Posted by Mary Jo Foley @ 6:01 pm

Categories: .Net Framework, Corporate strategy, Development tools, Multicore/distributed computing, Research, Visual Studio 10 ("Hawaii")

Tags: Memory, Microsoft Corp., Beta, .Net, Flash Memory, Application Servers, Middleware, Software Development, Software/Web Development, Enterprise Software

Microsoft released its first Beta of Visual Studio 2010 and the accompanying .Net Framework 4.0 back in May. On July 28, Microsoft announced it was releasing a second version of .Net 4.0 Beta 1.

Huh? That was my reaction when I saw a tweet about it earlier today. (Thanks for that, dotnetangel.)

But, yes, it’s true, as Senior Vice President of Microsoft’s Developer Division Soma Somasegar blogged today, calling the second version of .Net 4.0 Beta 1, STM.Net (with STM =  software transactional memory), an “experimental” release. Somasegar said Microsoft had made the STM.Net code available for download via the MSDN DevLabs site.

Somasegar explained more about the new release, which is designed with multi-core processing in mind, in his post:

“Transactional memory is a technology that frees developers from worrying about the mechanics of fine-grained locking and synchronization in multithreaded applications by providing transactional semantics for reading and writing to memory.  It enables developers to focus on application logic instead of the details of memory I/O when building multi-core and many-core programs.”

The DevLabs STM team added a few more details.

“This is an experimental release of the .NET Framework that allows C# programmers to try out this technology, specifically a particular implementation of STM. We are interested in your feedback on your experience using this programming model. Is it valuable and easy-to-use? Does it provide enough functionality? Are you willing to pay with serial performance losses to gain greater scalability? Our implementation is integrated with the framework and tools, it has been extended to provide coexistence with locks, interoperate with traditional transactional technologies, and safely work with existing code.”

Microsoft is expected to deliver the final Visual Studio 2010 and .Net 4.0 code in the spring of 2010.

July 17th, 2009

Microsoft releases Dryad concurrent-programming code to academics

Posted by Mary Jo Foley @ 7:09 am

Categories: .Net Framework, Azure, Code names, Corporate strategy, Development tools, Google, Multicore/distributed computing, Research, Utility/cloud computing, Windows server

Tags: Researcher, Microsoft Corp., Dryad, Dryad Software, DryadLINQ, Azure Piece, Aerospace & Defense, C#, Service Management, Cloud Computing

It’s been two years since Microsoft Chairman Bill Gates talked up Dryad, Microsoft’s concurrent-programming competitor to Google’s MapReduce and Apache Hadoop. But this week, Dryad was back on the radar screen, with Microsoft’s release of the Dryad code to academics and researchers.

The goal of Dryad is to enable programmers to develop and run applications on Windows computer clusters. The Dryad software is designed to allow systems to automatically parallelize at a low level complex applications across multiple machines.  Unlike existing high-performance computing and grid platforms, which are more focused on compute-intensive workloads, Dryad is more geared toward data-intensive computing scenarios where scale and fault-tolerance are of the essence.

Microsoft and a very few select partners have been using the Dryad code to develop a variety of sample and real-world apps — everything from bio-informatics to astronomy-focused programs.

At Microsoft’s Faculty Summit 2009 conference this week, company officials announced the availability of Dryad and the DryadLINQ programming tools. (The Dryad code is in binary form and DryadLINQ in source form). DryadLinq makes the Language-Integrated Query (LINQ) extensions to the C# programming language available to .Net developers writing Dryad apps. Researchers and academics can download both Dryad and DryadLINQ after signing Microsoft’s MS-Research licensing agreement.

At the conference, Microsoft execs also acknowleged that Dryad is not available on top of Azure, Microsoft’s cloud-computing platform, but that the plan is to make it available there. Officials didn’t offer Factulty Summit attendees a timetable for the Azure move, however.

(Check out the slide below from a March 2009 Dryad presentation by Microsoft Researcher Mihai Budiu. The Azure pieces are grayed out, but it’s possible to see how Microsoft envisions Dryad working with Azure.)

For now, Dryad is more research project than anything else. Like all Microsoft Research projects, there’s no guarantee if and when it will be commercialized.

But in 2008, Michael Isard, a Senior Researcher for Microsoft Research Silicon Valley, said Dryad “was built partly to help product groups with the short-term need they had to analyze data and partly in the hopes of being an enabling platform that would allow us to do research into other aspects of distributed computing.”

Microsoft has lots of different concurrent and distributed-computing projects at various stages of readiness. Midori, its next-generation operating system currently in incubation at the company, is just one of these.