Intel wants to thank all of the researchers who participated in the Many-Core Application Research Community. Intel's support for the Many-core Application Research Program ended in December, 2013. The information on this site is now read-only. There are no longer any active posts.
Intel announces Intel® Parallel Computing Centers
Intel is committed to making the journey to parallel computing together with you – beginning with a new program of Intel® Parallel Computing Centers. Intel Parallel Computing Centers are universities, institutions, and labs that are leaders in their field. The centers are focusing on modernizing applications to increase parallelism and scalability through optimizations that leverage cores, caches, threads, and vector capabilities of microprocessors and coprocessors. Code modernization is expected to enable large performance increases while maintaining the code portability users expect. Training engineers on modernizing their code is also necessary. To meet this need, the centers are developing a curriculum to equip students, scientists, partners and computer scientists with the skills to fully realize the capabilities of parallel computing resources. By enabling the advancement of parallelism, the Intel® Parallel Computing Centers will accelerate discovery in the fields of energy, finance, manufacturing, life sciences, weather, and beyond.
For more information go to http://software.intel.com/en-us/articles/intel-parallel-computing-centers. The Intel® Academic Program is accepting collaboration proposals until December 1st, 2013.
Many-Core Application Research Community Extended Through December 2013
The Many-Core Application Research Community includes more than 100 Institutions and 150 research projects with hundreds of participants world-wide. The MARC Program has exceeded Intel’s expectations with broad participation, software contributions and research results on the Single-chip Cloud Computer.
Software research has explored Barrelfish, X10, Bare Metal, Software Managed Coherence, Comm libraries, Message Passing Interfaces and more. The MARC Community has hosted numerous events and published nearly 100 papers on power and thermals, comms, memory management, programming languages, algorithms, and applications. More than a dozen institutions have active SCC-based curriculum to broaden Many-Core understanding and pursue research goals.
The MARC Program has built a world-wide research community and enhanced our understanding of many-core architecture and usage. Most MARC researchers have published papers or plan to do so soon. Intel has achieved a key many-core milestone with the announcement of Intel’s Xeon Phi™ product! Intel’s extending the MARC Program for a third year to help many-core researchers transition to the Intel® Xeon Phi™ product. Early in 2012 Intel closed the MARC Program to new participants. As announced in November 2012 at the MARC Symposium, Intel will end support for the MARC Program in December 2013. Research teams should focus on completing your research. As Intel works through the transition planning more details will be communicated.
Please see the April 2013 MARC Newsletter for more information or contact SCC_Research_Proposals@Intel.com
Based on interest from MARC community members, we are offering the option to have your name and contact information added to a MARC Researchers address book. This will only be available to registered MARC community members - not to the general public - through the MARC website. Please note that this is completely optional. We encourage interactions between researchers, this looks like a quick and easy way to make introductions.
Note: Access to address book is restricted to registered users only. Registration is now closed.
A special "Thank you" from The MARC Community to these individuals for their contributions to the MARC program.
· Jan-Arne Sobania of HPI has been a cornerstone of the MARC Community. His insights and carefully researched recommendations enabled other researchers to overcome problems and move forward quickly. Jan-Arne has been a key resource who resolved many complex technical challenges, and he has actively managed SCC Linux for the benefit of the MARC Community.
· Devendra Rai of ETH Zurich was one of the main beta testers of our 22.214.171.124 sccKit and our latest sccLinux. He is very active in the community. Since the last newsletter, he has authored 7 new posts, participated in 11 and reported 12 new bugs. Devendra has been a great catalyst for the community. Thank you!
Intel recognizes MARC members who make the extra effort to grow and enrich the community. As a MARC community member, do you know of someone or some group you would like to nominate for Intel’s MARC recognition? If you do, please send mail to SCC_Research_Proposals@intel.com. Please specify the name of the person or group and give a short (a sentence or two) description of why you are making the nomination.
|Default memory consistency model of SCC 4 months ago||by Nil|
|Re: Relation between CPU freq and Memory freq 5 months ago||by chan seok kang|
|How does sccLinux knows which frequency it is running? 5 months ago||by Nil|
|How to check data in shared memory. 6 months ago||by chan seok kang|
|Question when creating shared memory on SCC by means of LUT remapping. 6 months ago||by Nil|
|Changing voltage directly on SCC does not work as expected 6 months ago||by Nil|
|Re: Strange, but regular, long-term power fluctuations on SCC 9 months ago||by mwvantol|
|Re: MARC Lab Data Center connectivity down 10 months ago||by mwaughex|