Intel® Software Network Blogs » Aaron Tersteeg (Intel)

Research@Intel Day - Go!

Aaron Tersteeg (Intel) — Thu, 12 Jun 2008 20:42:28 +0000

Yesterday's Research @ Intel Day was a huge success. Lots of media participation, amazing demonstrations of the great research Intel is cooking up in the labs and a two very important guests.

The day started out with a key note presentation by Justin Ratner, Intel CTO and Fellow. He covered the highlights of what themes were addressed at the event. Green products and energy effecency were very hot topics along with Visual Computing and Tera-Scaling. The WhatIf team presented the verity of experimental projects available for free download and demonstrated three key projects of interest to Parallel Software Developers.

Intel® Adaptive Spike-Based Solver - a highly scalable and adapted solver for large banded systems presented by Peter Tang
Intel® C++ STM Compiler, Prototype Edition 2.0 - a new approach to memory management presented by Ravi Narayanaswamy
Intel® Performance Tuning Utility 3.0 -a powerful, cross-platform performance analysis tool set for optimizing code presented by Bhanu Shankar

We met with many interested press people including Jon Fortt from Fortune, Kevin Rockwell from CNN, Darold Massaro from PodTech, Ann Mutschler for EDN, Gote Fagerfjall from Swedish Press, Ben Flanagan from UAE Press, Elsa Wenzel from CNET and our very own Geek Blogger Josh Bancroft.

Jim Dempsey, a frequent contributor to the Parallel Programming Forum, joined us for the event an seemed to enjoy all the innovations. It was great to to meet Jim for the first time and we all look forward to his continued high level of involvement in the Multi-Core community.

Another guest that we were delighted to talk to was Paul Otellini, Intel CEO. Who stopped by our booth to discuss the WhatIf site with Bhanu and the current status of Software Transactional Memory with Ravi. The guys did an amazing job telling the story again and again to all our enthusiastic guests.

It was a long day full of learning new things, sharing our ideas and connecting with all the bright people who do Research@Intel. Read more about the event on Ken Kaplan's blog.

Research@Intel Day - Ready, Set...

Aaron Tersteeg (Intel) — Wed, 11 Jun 2008 01:20:15 +0000

Busy day today coordinating the set up of our gear and getting the demos installed. I'm super excited about the brand new ALIENWARE AREA 51 3.0 QUAD CORE machine we got hooked up with. I wish that this was the machine that I have in my home office. We had a few last minute changes (we now have wired internet access and our poster is going to be updated). We have a great video for Peter Tang to showcase the benefits of the Intel Adaptive Spike Based Solver. Ravi Narayanaswamy is ready with a quick and to the point demo the Intel C++ Software Transactional Memory Compiler. Bhanu Shankar has installed the newest version of the Intel Performance Tuning Utility and is excited about showing what it can do. And I the ring leader tying it all together under the Multi-Core projects in the WhatIf Prototype Software Site. More links to What If projects.

Parallel Programing Tools from WhatIf.intel.com showcased at Research@Intel Press Event

Aaron Tersteeg (Intel) — Tue, 10 Jun 2008 06:14:34 +0000

Wednesday June 11th is Research@Intel Day 2008 - Intel's once a year invitation only press and educational community event designed to showcase the great ideas that the Corporate Technology Group and and other research groups are developing. We have been working on putting together the event for months and I've very excited to get out and show off some of our great work. I'm staffing a booth with 3 other engineers to talk about the WhatIf Site and show cases 3 of the parallel programing innovations from Intel's Software Research Lab.

Intel® Adaptive Spike-Based Solver - a highly scaleable and adapted solver for large banded systems presented by Peter Tang
Intel® C++ STM Compiler, Prototype Edition 2.0 - a new approach to memory management presented by Ravi Narayanaswamy
Intel® Performance Tuning Utility 3.0 - a powerful, cross-platform performance analysis tool set for optimizing code presented by Bhanu Shankar

If your in the press and read my blog then please stop by and say hello. If your not in the press then you'll just have to wait a few hours for the more more savvy media folks to get the word out.

Parallel Programing Glossary of Technical Terms

Aaron Tersteeg (Intel) — Tue, 27 May 2008 22:38:41 +0000

The Parallel Programing Glossary of Technical Terms has been published in the Intel Software Network Wiki. We've started with a list of terms we use often and look forward to your edits and update.

Thank you to Tim Mattson and his co-authors Beverly A. Sanders, Berna L. Massingill for allowing us to publish the terms from their book Patterns for Parallel Programming.

Here is is the list of terms, if you see something missing please join us in editing to make the list the best it can be:

A Community for Financial Services Software Development

Aaron Tersteeg (Intel) — Wed, 02 Apr 2008 21:44:48 +0000

It has been too long since my last blog post and I'm sure that your wondering what I've been upto. I have a few project in the works and the first that I'd like to share is the new Financial Services Community that was launched on March 28, 2008. In late 2007 Ajay Mungara starting working with the engineering team that serves the Financial Services Industry to identify a community based approach to share their engineering lessons learned with the online community of software developers. After doing some research it was determined that the developer were interested in Manageability but passionate about performance. Speed was what they needed and optimization for Multi-Core seemed to be a better initial focus for the community. Our peers in Europe has already started to market directly to the business people in Financial Services through the FasterFS web site with the goal of sharing the success of Intel's low-energy consumption and lightning performance but the site does note really address the "hot to" that many developer want.

A few months ago I started working with Sunil, Shuo and George to put together a content plan, identify the right enabling engineers, get some blog posts going and set up a forum to address developer questions. The site is now live with our first batch of content and ideas. If you work for a financial services company or write software for the industruy then jump right into the conversation. Let us know what you like and where we can help. We look forward to your questions and comments.

CFinder Capabilities, Results and Analysis - What level of parallelism does your application have?

Aaron Tersteeg (Intel) — Tue, 19 Feb 2008 04:55:54 +0000

Khang Nguyen, Intel Application Engineer in the Software & Solutions Group, sat down with me a few weeks ago to talk about CFinder. The first video I posted was an overview of the key features and benefits of the concurrency audit tool. As a follow up we are presenting two more videos.

In the second video he reviews the capabilities and usage of CFinder for detecting parallelism in an application. He starts with the Main Screen that presents processor information including sockets, cores, logical processors and other processor features. He then defines the process to define the application that CFinder collects data for, the run time duration and sampling rate. After CFinder is run it present the user with CPU utilization, elapsed time, parallel time and concurrent levels.

In the third video Khang presents result data from CFinder and discusses how to interpret the results. Two key indicators from CFinder are Parallel level, the time multiple threads running and Concurrent level, the number of threads and the percentage of time for each threads. Understanding what each number means and how to interoperate the results are key to understanding the potential for multi-core optimization.

Video: Benefit of Using CFinder

Aaron Tersteeg (Intel) — Thu, 07 Feb 2008 19:04:49 +0000

I asked Khang Nguyen, Intel Application Engineer in the Software & Solutions Group to do a brief video to Introduces you to CFinder, the tool for detecting parallelism in an application. In his talk he covers the business perspective for developing mulit-threaded applications, the issues the developers face, the key feature and benefits of Cfinder. Check out the video, download the free tool and let us know what you think.

Using Active Management Technology And Virtualization To Manage Application On Client Platforms.

Aaron Tersteeg (Intel) — Fri, 25 Jan 2008 17:50:02 +0000

At Fall IDF 2007 Ylian Saint-Hilaire and I talked with Brian Duckering of AppStream about his company and how Intel’s AMT and Virtualization can be used for better application management.

AppStream provides application streaming for efficient distribution of applications to systems in the enterprise. It facilities a centralized process for licensing, managing and updating the deployment of application on demand. Users can use any machine, login, and instantly have access to all the applications that are available to them.

AppStream uses AMT heartbeat detection to identify systems state as well as third party data store to manage and track licensees and data on application availability to user.
With most licensing models it is difficult to know fill license state when some machine are powered down. When the machine is off it is an empty spot on the network.

AppStream's IDF demonstration allows for a console view to read the Third Party Data Store from AMT and gives the user application list on the machine. If a license is unused it can be transferred even if the machine is off. On wake up it reads the client can read the Third Party Data Store and make update to the system based on availability and utilization of licensing. This increases the accountability and efficient utilization of available application licenses.

AppStream can also utilize virtualization to a managed layer that is controlled through a filter driver. Multiple application versions can be run on the same system through virtualization to allow for coexisting multiple versions of the same dll on the same system. Read more about AppStream at http://www.appstream.com/

Video of Clay Breshears, Intel Software College Courseware Architect

Aaron Tersteeg (Intel) — Mon, 07 Jan 2008 15:04:56 +0000

Many of you have probably been reading Clay Breshears blog on Parallel Programming. We thought that you would be interested in learning more about Clay’s role within Intel and what he spends most of his time doing when he is not blogging on the Intel Software Network. During a recent conference we both sat down to discuss the Intel Software College and his contribution to parallel software education.

Clay explains that Intel offers course directly to customers and they work with third party training companies to ensure that the global market can get the training it needs. The current focus of the Intel Software College is on Multi-Core and Threading introductory, intermediate and advanced level courses. They provide training on Windows Threads, POSIX Threads, OpenMP, Intel® Threading Building Blocks and the use of Intel Software Tools (Intel® Thread Checker, Intel® Thread Profiler, Intel® VTune™ Performance Analyzer, the Intel® Compilers).

A typical class would include both lecture with discussion followed by a hands on lab to apply the techniques discussed. Advanced classes include the use of Intel tools to address race conditions and dead locks. The language of choice for most classes is C and C++. Intel is currently offering introductory courses for Java as well. Intel has extending the programs into University and Colleges through Intel’s Higher Education Program.

Software Engineers focus is on foundational skills but University Professors are much more interested in the more advanced curriculum. They realize that scalability of software is the next big thing as the market goes from 2 to 4 to even more processor on a chip. The University programs are changing their curriculum to address the mainstream adoptions of multi-core systems.

Threading for Multi-Core Highlights - Update from the Community Manager

Aaron Tersteeg (Intel) — Thu, 20 Dec 2007 23:02:52 +0000

The Treading for Multi-Core community has had a great 2007! Thank you to everyone that contributed content, opinion, comments, contest entries and votes. We are looking forward to exciting new technology and even more engaging discussions in 2008. Here are a few highlights from the past month or so, I’ll try to recap some of 2007 in a follow up post.

The mid November release of Penryn was a great success for Intel. The new SSE4 instructions allow developers to create amazing fast video encoding applications. If you have not seen Penryn in action I would strongly suggest that you watch the Penryn demo video that Darren Yee and I made at Fall IDF. The speed up is amazing. If you are wondering how fast the new chips are then I would suggest you check out Intel's new chip performance page for the latest benchmarking data.

The community and even some folks at Intel have asked for a central place to keep track of all the information, tools, programs, media and events around Intel’s Multi-Core and Multi-Processor platforms. We have heard your request and are excited with our new Ready, Thread, Go! section of the Threading for Multi-Core Community. It is your technical training ground complete with all the tools, resources and expertise you need to thread your applications and solve your technical challenges. Check out each section and get the most out of the Intel Threading for Multi-Core Community.

The Parallel Programing forum continue to see an amazing amount of traffic. In the month of December we had 15 active discussion going on in addition to all the community members viewing our 431 discussion topics.

One of the most interesting threads for me has been the discussion going on about simulating multi-core systems. You should check it out and share your thoughts.

University Professors have joined the community. Please welcome Prof. Sam Midkiff and Prof. Matthew Wolf to our University Curriculum Blog. They have been sharing their thoughts on including parallel programming and multi-core software development in the undergraduate curriculum.

Last month the Intel Software Network team hit the road in Europe and the Middle East with the “On the Road to Concurrency” tour. If you were not able to attend the events we have posted the presentation from Düsseldorf, Ankara and Moscow as well as the presentation from the event in Israel. We have also added footage from the events in the Community Uploads sections of our videos.

If you have not tried out CFinder then you should. CFinder is Intel's new Free Application Concurrency Audit Tool.