I am a Postdoctoral Research Associate at Carnegie Mellon University (CMU) in Qatar. I have a broad interest in computer systems with an emphasis on cloud computing and computer architecture. For my Ph.D. thesis, I focused on L2 cache design of multicore processors. After joining CMU in 2011, I extended my work to cloud computing where I devised multiple MapReduce scheduling techniques and optimized task parallelism for improved Hadoop performance. I hold a BS in Computer Science from the American University of Science and Technology, Lebanon, as well as MS and PhD in Computer Science from the University of Pittsburgh, USA.

Teaching

At Carnegie Mellon University (CMU) in Qatar:

• 15-346- Perspectives in Computer Architecture, Spring 2013
• 15-319- Cloud Computing, Spring 2013
• 15-440- Distributed Systems, Fall 2012
• 15-319- Cloud Computing, Spring 2012
• 15-440- Distributed Systems, Fall 2011

Prior to Joining CMUQ:

• CSI 412- Advanced Computer Architecture, Fall 2010 at AUST
• CSI 311- Java Programming, Fall 2010 at AUST
• CSI 250- C++ Programming, Fall 2010 at AUST
• CS110- Introduction to Personal Computers and the Internet, Fall 2007 at the University of Pittsburgh
• CS 449- Introduction to Systems Software, Fall 2006 at the University of Pittsburgh (Lab Instructor and TA)
• CS/COE 1550- Introduction to Operating Systems, Summer 2006 at the University of Pittsburgh (Lab Instructor and TA)

Research Projects

Scheduling in Hadoop:

The dynamic nature of cloud infrastructure requires significant advancements in application workflow management and scheduling. In this research I am working towards improving scheduling mechanisms for applications running on the cloud. Specifically, I am working on applying data locality to Reduce task scheduling in Hadoop, thereby reducing network traffic and enhancing the performance of MapReduce jobs. I am also working on new monitoring and scheduling techniques to better detect faulty and slow tasks in MapReduce and intelligently alleviate their negative effects on the overall system performance. [CloudCom2011, CLOUD2012, Two Book Chapters on Virtualization and Data Analytics for the Cloud, Proposal Accepted, CRC Press]

Workload Characterization:

In a dynamic infrastructure environment such as clouds, it is critical to understand an application’s resource needs and behavior.  This enables intelligent configuration of the incubating distributed analytics engines as well as effective provisioning of virtual resources that are tied to the needs of applications, among others. In this research I aim to characterize various scientific applications to influence provisioning, configuration as well as static and dynamic optimization for such applications on cloud computing systems. [In-Progress. Paper Submitted to CLOUD2013]

Cloud Monitoring:

Deploying performance-sensitive applications on the cloud is cumbersome due to the complexity of the application execution environment. Routine tasks such as monitoring, performance analysis and debugging often require close interaction and inspection of multiple layers in the application and system software stack. In this research I aim to design monitoring tools that help integrate metrics into these layers and allow for flexible visualization and analysis. [CloudCom2011]

CMP Cache Management:

As large uniprocessors are no longer scaling in performance, CMPs have become the trend in computer architecture. CMPs can easily spread multiple threads of execution across various cores. Besides, CMPs scale across generations of silicon process simply by stamping down copies of the hard-to-design cores on successive chip generations. A key requirement to obtaining high performance from CMPs is to effectively manage the limited on-chip cache resources. In this research I present a general framework for approaching cache management in CMPs and aim to explore novel CMP cache designs that effectively employ the framework and efficiently achieve scalable performance.[HiPEAC 2009, ICS 2009, CAL 2010, PhD Dissertation 2010, PACT 2010, HiPEAC 2011, JPDC 2011, Book Chapter Submitted to CRC Press]

High-Performance Memory Substrates for Search-Intensive Applications:

Search operations can occupy a significant portion of total execution time and energy consumption, while posing a difficult performance problem to tackle using traditional memory hierarchy concepts. In this research I aim to extend the conventional content addressable memory to accelerate search operations present in many important real-world applications (e.g., IP address lookup in core routers and trigram lookup in large speech recognition systems). [ISPASS 2007]

Power-Aware Memory Management using Software Generated Hints:

Current state-of-the-art power-aware DRAM chips offer various power modes (active, standby, nap, and powerdown) in order to provide a potential to limit power consumption in the face of increasing demand for performance. In response to workloads becoming increasingly memory-intensive and data-centric, in this research I aim to utilize and exploit various power modes for the most effective main memory power management. [Technical Report TR-09-163- Provided to Intel (Proprietary of Intel)]

Publications

• Mohammad Hammoud and Majd F. Sakr, "Virtualizing Resources for the Cloud" (To appear) Advances in Data Processing Techniques in the Era of Big Data, CRC Press, 2013.

• Mohammad Hammoud and Majd F. Sakr, "MC²: Map Concurrency Characterization for MapReduce on the Cloud" In Proceedings of The Sixth International Conference on Cloud Computing (CLOUD 2013), Santa Clara, California, USA, June 2013, (To Appear).

• Mohammad Hammoud, M. Suhail Rehman and Majd F. Sakr, "Center-of-Gravity Reduce Task Scheduling to Lower MapReduce Network Traffic" In Proceedings of The Fifth International Conference on Cloud Computing (CLOUD 2012), Honolulu, Hawaii, USA, June 2012 [PDF].

• Mohammad Hammoud and Majd F. Sakr, "Locality-Aware Reduce Task Scheduling for MapReduce" In Proceedings of  The 3rd International Conference on Cloud Computing and Science (CloudCom 2011), Athens, Greece, December 2011 [PDF].

• M. Suhail Rehman, Mohammad Hammoud, Majd F. Sakr, "VOtus: A Flexible And Scalable Monitoring Framework for Virtualized Clusters" (Poster Paper), In Proceedings of The 3rd International Conference on Cloud Computing and Science (CloudCom 2011), Athens, Greece, December 2011 [PDF].

• Mohammad Hammoud and Majd F. Sakr, "Task Scheduling in MapReduce: Principles, Challenges, Mechanisms and Future Directions" (Book Chapter- Proposal Submitted to CRC Press).

• Mohammad Hammoud, Sangyeun Cho, and Rami Melhem, "FSB: Flexible Set Balancing Strategy for Last Level Caches" (Book Chapter- Submitted to CRC Press).

• Mohammad Hammoud, Sangyeun Cho, and Rami Melhem, C-AMTE: A Location Mechanism for Flexible Cache Management in Chip Multiprocessors. Journal of parallel and Distributed Computing (JPDC), June 2011 [PDF].

• Mohammad Hammoud, Sangyeun Cho, and Rami Melhem, Cache Equalizer: A Placement Mechanism for Chip Multiprocessor Distributed Shared Caches. Proceedings of the 6th Int'l Conference on High Performance and Embedded Architectures and Compilers (HiPEAC), Heraklion, Crete, Greece, January 2011 [PDF].

• Mohammad Hammoud, Sangyeun Cho, and Rami Melhem, An Intra-Tile Cache Set Balancing Scheme. (Poster Paper), Proceedings of the Int'l Conference on Parallel Architectures and Compilation Techniques (PACT), Vienna, Austria, September 2010 [PDF].

• Mohammad Hammoud, Hardware-Oriented Cache Management for Large-Scale Chip Multiprocessors. PhD Dissertation, August 2010 [PDF].

• Mohammad Hammoud, Sangyeun Cho, and Rami Melhem, A Dynamic Pressure-Aware Associative Placement Strategy for Large Scale Chip Multiprocessors. IEEE Computer Architecture Letters (CAL), January-June 2010 [PDF].

• Mohammad Hammoud, Sangyeun Cho, and Rami Melhem, Dynamic Cache Clustering for Chip Multiprocessors. Proceedings of the ACM Int'l Conference on Supercomputing (ICS), IBM T. J. Watson Research Center, New York, June 2009 [PDF].

• Mohammad Hammoud and Rami Melhem, Exploratory Efforts to Manage Power-Aware Memories using Software Generated Hints. Technical Report TR-09-163, Department of Computer Science, University of Pittsburgh (Proprietary of Intel).

• Mohammad Hammoud, Sangyeun Cho, and Rami Melhem, ACM: An Efficient Approach for Managing Shared Caches in Chip Multiprocessors. In Proceedings of the 4th Int'l Conference on High Performance and Embedded Architectures and Compilers (HiPEAC) , Paphos, Cyprus, January 2009 [PDF].

• Sangyeun Cho, Joel R. Martin, Ruibin Xu, Mohammad H. Hammoud and Rami Melhem. CA-RAM: A High-Performance Memory Substrate for Search-Intensive Applications. In IEEE Int'l Symposium on Performance Analysis of Systems and Software (ISPASS), San Jose, California, April 2007 [PDF].