SEARCH
ISTC-CC NEWSLETTER
RESEARCH HIGHLIGHTS
Ling Liu's SC13 paper "Large Graph Processing Without the Overhead" featured by HPCwire.
ISTC-CC provides a listing of useful benchmarks for cloud computing.
Another list highlighting Open Source Software Releases.
Second GraphLab workshop should be even bigger than the first! GraphLab is a new programming framework for graph-style data analytics.
ISTC–CC Q4 2013 Status Report
Summary:
The ISTC-CC held its third annual retreat in Pittsburgh on November 7 & 8. The 106 attendees included faculty and students from Carnegie Mellon, Georgia Tech, Princeton, UC Berkeley, and Washington, as well as 21 Intel employees. The agenda featured keynotes by Rich Uhlig, Pradeep Dubey, and Myles Wilde of Intel, 13 research talks by faculty and students from all five Universities, 4 BoF sessions, and 44 posters. By all accounts, the retreat was a big success: great interactions, lots of connections made, new insights, idea inspiration, and generally superb energy! Further information and slides can be found at http://www.istc-cc.cmu.edu/events/retreat13.shtml (user=istc-cc, pw=altostratus).
The retreat was followed by the Board of Advisors meeting where Greg Ganger and Phil Gibbons presented the ISTC-CC’s plans for years 3-5. Phil Gibbons and Jeff Parkhurst presented additional Intel-only material for a meeting with Intel stakeholders and Intel BoA members. These meetings provided considerable positive feedback, as well as good suggestions. In early December we learned that the ISTC-CC was renewed for years 4-5 at the $2M/year level.
For further information on the center’s research, please see the ISTC-CC 2013 Newsletter, which highlights the research, publications, and news of the past 5 quarters. It is available at www.istc-cc.cmu.edu/publications/news/ISTC-newsletter13.pdf. Jeff Parkhurst is working with James Gurganus (Intel Labs) to set up an internal mirroring site to archive code that is generated from the center research.
New IEEE Fellows: Congratulations to Phillip Gibbons (Intel Labs), Garth Gibson (CMU) and Sudhakar Yalamanchili (GA Tech) were named IEEE Fellows, bringing ISTC-CC’s totals to 8 IEEE Fellows and 8 ACM Fellows.
Kudos also to:
- Dan Siewiorek (CMU), for ACM/IEEE Design Automation Conference's Second Decade (1974-1983) Award, Top 10 Author Award, and Prolific Author Award
- Dong Zhou, Bin Fan, Hyeontaek Lim (CMU grad students), Michael Kaminsky (Intel Labs), and David Andersen (CMU), for their paper on fast switching based on cuckoo hashing being one of four papers nominated for Best Paper at CoNEXT’13
- Ion Stoica (UC Berkeley), for serving as NSDI’14 Program Co-Chair; Mike Kozuch (Intel Labs), selected as HotCloud’14 Program Co-Chair; and Karsten Schwan (GA Tech), for serving as Middleware’13 Program Co-Chair
- Michael Kaminsky and Babu Pillai (Intel Labs), for organizing a highly successful SOSP’13 conference
Publications: A strong quarter for publications, with 30 papers published (including 4 SOSP’13 papers and 3 SOCC’13 papers) and another 13 papers accepted for publication!
Amplifying Funding:
- Justin Meza and Lavanya Subramanian (CMU grad students) were each awarded CMU College of Engineering Bertucci Fellowships.
- Onur Mutlu’s (CMU) NSF CNS project on "single-level stores" was funded.
Intel Interactions Highlights:
- ISTC-CC research was highlighted in meetings with BU stakeholders Balint Fleischer (DCG), Jonathan Donaldson (DCG), Eric Dahlen (DCG), Girish Juneja (SSG/DSD), Vin Sharma (SSG/DSD), Avik Dey (SSG/DSD), and Geoff Lowney (SSG), among others, providing highly positive stakeholder feedback on the value of the research projects, as well as suggestions for furthering the value propositions.
ISTC-CC “does phenomenal work – some of it goes directly into products.” – Girish Juneja, CTO, Datacenter Software Division
Technical highlights
- Publications summary for the past quarter:
- 13 papers accepted: to HPCA'14 (4 papers), FAST'14 (2 papers), NSDI'14 (2 papers), ASPLOS'14, CGO'14, HotMobile'14, IC2E'14, VLDB'14
- 30 papers published: to SOSP'13 (4 papers), SOCC'13 (3 papers), ICCD'13 (2 papers), MICRO'13 (2 papers), TRIOS'13 (2 papers), BigData'13, BigLearning'13, Cluster'13, CoNEXT'13, FOCS'13, ICDM'13, INFLOW'13, InfoVis'13, NIPS'13, OPT'13, PerCom'13, SC'13, DAPD, Geoinformatica, IEEE TMC, OSR, SUSCOM
- Open Source code releases this past quarter included:
- Compressed Buffer Tree (CBT) source code release: https://bitbucket.org/sopwithcamel/cbt. The system itself is described in Amur et al.'s SOCC 2013 paper. (Co-authored by GA Tech, CMU, Intel)
- Alex Smola (CMU) along with Dave Andersen, Mu Li, Li Zhou, Aaron Li, and Fei Xia (CMU) released an alpha release of their distributed, fault tolerant and replicating parameter server. It allows intelligent updates on the server end and for easy parallelization of existing codes using this framework.
- Georgia Tech’s research milestones for the past quarter were determined by PhD student graduations:
- Min Lee (former GA Tech grad student now at Intel) completed his PhD thesis, entitled "`Memory Region: A System Abstraction for Managing the Complex Memory Structures of Multicore Platforms,'' November 2013. Committee members included Rob F. Van der Wijngaart (Intel).
- Hrishikesh Amur (PhD), “Storage and Aggregation for Fast Analytics Systems,'' October 2013, committee members included Dave Andersen (CMU), Greg Ganger (CMU) and Michael Kaminsky (Intel Labs).
- Chengwei Wang (PhD), “Monitoring and Analysis System for Performance Troubleshooting in Data Centers,” October 2013.
- Vishal Gupta (PhD), “Systems Abstractions for Resource Scaling on Heterogeneous Platforms,” September 2013, committee members included Scott Hahn (Intel Labs) and George Cox (Intel).
These students are working on further papers from their thesis work. Congrats to the students on completing their PhDs!
- “Efficient Data Partitioning Model for Heterogeneous Graphs in the Cloud,” Kisung Lee and Ling Liu, was selected by HPCWire as a featured article from the SC 2013 proceedings <http://www.hpcwire.com/2013/11/16/sc13-research-highlight-large-graph-processing-without-overhead/#/>
Sponsor group interaction highlights
- Onur Mutlu (CMU) co-authored two papers with Zehan Chishti, Samira Khan, Alaa Alameldeen, and Chris Wilkerson (all Intel Labs) that were accepted to HPCA’14.
- Matt Wolf (GA Tech) has been interacting with Intel's Michael McCool and others to define a Georgia Tech hackfest (planned for Feb. 2014) around Intel's new Internet of Things toolkit with Quark processors.
- Ada Gavrilovska and Karsten Schwan (GA Tech) had discussions with Intel's Rob Knauerhase and separately, with Intel's Shekhar Borkar at Supercomputing 2013.
- Karsten Schwan interacted with Intel's Vishakha Gupta and Rich Uhlig at SOSP 2013.
- Intel's Vishakha Gupta and Romain Cledat visited Georgia Tech in December 2013.
- Myles Wilde (DCG) and Jay Kyathsandra (DCG), facilitated by Mike Kozuch (Intel Labs), came to CMU in October to explore opportunities for the Quark/Galileo platform with Professors Yuvraj Agarwal, Priya Narasimhan, Anthony Rowe, Dave Andersen, and Majd Sakr.
- Phil Gibbons hosted a visit by Jeff Smits (SSG), who met with Greg Ganger, Garth Gibson, Onur Mutlu, and Andy Pavlo from CMU, and Michael Kaminsky, Mike Kozuch, and Babu Pillai from Intel, to discuss NVM systems research. Phil and Jeff had a follow-up meeting to plan next steps.
- Michael Kaminsky, David Andersen (CMU), and Dong Zhou (CMU grad student) continue to meet with Ren Wang (Intel Labs) and Maziar Manesh (Intel) about their CuckooSwitch work. The ISTC team provided them a copy of their code and their CoNEXT’13 paper.
Other ISTC highlights
- 2014 IEEE Fellows included:
- Phil Gibbons (Intel Labs), for contributions to parallel computing and databases
- Garth Gibson (CMU), for contributions to the performance and reliability of transformative storage systems
- Sudhakar Yalamanchili (GA Tech), for contributions to high-performance multiprocessor architecture and communication
- [Amplifying Funding] A proposal by Mahadev Satyanarayanan (PI) and co-PIs, Dan Siewiorek, Jason Hong, and Asim Smailagic (CMU), entitled, “QuiltView: Glass-Sourced Video for Google Maps Queries” was submitted to Google and selected for receiving ten copies of Google Glass. Also, Iulian Moraru won an (unsolicited) Facebook Research Collaboration Award, which includes one year of funding. Unfortunately (fortunately :), Iulian is graduating and had to decline.
List of publications
- “Core Groups: System Abstractions for Extending the Dynamic Range of Client Devices using Heterogeneous Cores,” Vishal Gupta, Paul Brett, David Koufaty, Dheeraj Reddy, Scott Hahn, Karsten Schwan, and Ganapati Srinivasa, Elsevier Journal of Sustainable Computing (SUSCOM), 3(3), selected papers from the 2012 IEEE International Green Computing Conference, September 2013.
- “Road-Network Aware Trajectory Clustering: Integrating Locality, Flow and Density,” Binh Han and Ling Liu, IEEE Transactions on Mobile Computing, vol. 12, September 2013.
- “Oncilla: A GAS Runtime for Efficient Resource Allocation and Data Movement in Accelerated Clusters,” J. Young, S. H. Shon, S. Yalamanchili, A. Merrit, K. Schwan and H. Froening, IEEE International Conference on Cluster Computing (Cluster’13), September 2013.
- “Visualizing Request-flow Comparison to Aid Performance Diagnosis in Distributed Systems,” Raja R. Sambasivan, Ilari Shafer, Michelle L. Mazurek, and Gregory R. Ganger, InfoVis’13, October 2013.
- “The Role of Cloudlets in Hostile Environments,” M. Satyanarayanan, G. Lewis, E. Morris, S. Simanta, J. Boleng, and K. Ha, IEEE Pervasive Computing (PerCom’13), October 2013.
- “Program Interference in MLC NAND Flash Memory: Characterization, Modeling, and Mitigation,” Yu Cai, Onur Mutlu, Erich F. Haratsch, and Ken Mai, 31st IEEE International Conference on Computer Design (ICCD’13), October 2013.
- “LightTx: A Lightweight Transactional Design in Flash-based SSDs to Support Flexible Transactions,” Youyou Lu, Jiwu Shu, Jia Guo, Shuai Li, and Onur Mutlu, 31st IEEE International Conference on Computer Design (ICCD’13), October 2013.
- “Memory-Efficient GroupBy-Aggregate using Compressed Buffer Trees,” Hrishikesh Amur, Wolfgang Richter, David G. Andersen, Michael Kaminsky, Karsten Schwan, Athula Balanachandran, and Erik Zawadzki, 4th ACM Symposium on Cloud Computing (SOCC’13), October 2013.
- “vTube: Efficient Streaming of Virtual Appliances Over Last-Mile Networks,” Yoshihisa Abe, Roxana Geambasu, Kaustubh Joshi, H. Andres Lagar-Cavilla, and Mahadev Satyanarayanan, 4th ACM Symposium on Cloud Computing (SOCC’13), October 2013.
- “Hierarchical Scheduling for Diverse Datacenter Workloads,” Arka Bhattacharya, Eric Friedman, Ali Ghodsi, Scott Shenker, and Ion Stoica, 4th ACM Symposium on Cloud Computing (SOCC’13), October 2013.
- “An Infrastructure for Automating Large-scale Performance Studies and Data Processing,” Deepal Jayasinghe, Josh Kimball, Tao Zhu, Siddharth Choudhary, and Calton Pu, IEEE Big Data Conference (IEEE BigData’13), October 2013.
- “Iterative Row Sampling,” Mu Li, Gary L. Miller, and Richard Peng, 54th Annual IEEE Symposium on Foundations of Computer Science (FOCS’13), October 2013.
- “Performance Troubleshooting in Datacenters,” Chengwei Wang, Soila Pertet Kavulya, Jiaqi Tan, Michael Kasick, Liting Hu, Mahendra Kutare, Priya Narasimham, Karsten Schwan, and Rajeev Gandhi, Operating Systems Review, October 2013.
- “Consistent, Durable, and Safe Memory Management for Byte-addressable Non Volatile Main Memory,” Iulian Moraru, David G. Andersen, Michael Kaminsky, Parthasarathy Ranganathan, Niraj Tolia, and Nathan Binkert, First ACM Conference on Timely Results in Operating Systems (TRIOS’13), with SOSP’13, November 2013.
- “inTune: Coordinating Multicore Islands to Achieve Global Policy Objectives,” Priyanka Tembey, Ada Gavriloska, and Karsten Schwan, First ACM Conference on Timely Results in Operating Systems (TRIOS’13), with SOSP’13, November 2013.
- “NVM Heaps for Accelerating Browser-based Applications,” Sudarsan Kannan, Ada Gavrilvoska, and Karsten Schwan, Workshop on Interactions of NVM/Flash with Operating-Systems and Workloads (INFLOW’13), with SOSP’13, November 2013.
- “Efficient Data Partitioning Model for Heterogeneous Graphs in the Cloud,” Kisung Lee and Ling Liu, IEEE international Conference for High Performance Computing, Networking, Storage and Analysis (SC2013), November 2013.
- “Sparrow: Distributed, Low Latency Scheduling,” Kay Ousterhout, Patrick Wendell, Matei Zaharia, and Ion Stoica, 24th ACM Symposium on Operating Systems Principles (SOSP'13), November 2013.
- “There Is More Consensus in Egalitarian Parliaments,” Iulian Moraru, David G. Andersen, and Michael Kaminsky, 24th ACM Symposium on Operating Systems Principles (SOSP'13), November 2013.
- “PARROT: A Practical Runtime for Deterministic, Stable, and Reliable Threads,” Heming Cui, Jiri Simsa, Yi-Hong Lin, Hao Li, Ben Blum, Xinan Xu, Junfeng Yang, Garth A. Gibson, and Randal E. Bryant, 24th ACM Symposium on Operating Systems Principles (SOSP'13), November 2013.
- “Discretized Streams: Fault-Tolerant Streaming Computation at Scale,” Matei Zaharia, Tathagata Das, Haoyuan Li, Timothy Hunter, Scott Shenker, and Ion Stoica, 24th ACM Symposium on Operating Systems Principles (SOSP'13), November 2013.
- “More Effective Distributed ML via a Stale Synchronous Parallel Parameter Server,” Q. Ho, J. Cipar, H. Cui, S. Lee, J. Kim, P. Gibbons, G. Gibson, G. Ganger, and E. Xing, Neural Information Processing Systems Conference (NIPS’13), December 2013.
- “Distributed Delayed Proximal Gradient Methods,” Mu Li, Dave Andersen, and Alex Smola, Workshop on OPT2013: Optimization for Machine Learning, with NIPS’13, December 2013.
- Mu Li, Li Zhou, Zichao Yang, Aaron Li, Fei Xia, David G. Andersen, and Alexander Smola, “Parameter Server for Distributed Machine Learning,” Workshop on Big Learning : Advances in Algorithms and Data Management, with NIPS’13, December 2013.
- “RowClone: Fast and Energy-Efficient In-DRAM Bulk Data Copy and Initialization,” Vivek Seshadri, Yoongu Kim, Chris Fallin, Donghyuk Lee, Rachata Ausavarungnirun, Gennady Pekhimenko, Yixin Luo, Onur Mutlu, Phillip B. Gibbons, Michael A. Kozuch, and Todd C. Mowry, 46th IEEE/ACM International Symposium on Microarchitecture (MICRO-46), December 2013.
- “Linearly Compressed Pages: A Main Memory Compression Framework with Low Complexity and Low Latency,” Gennady Pekhimenko, Vivek Seshadri, Yoongu Kim, Hongyi Xin, Onur Mutlu, Phillip B. Gibbons, Michael A. Kozuch, and Todd C. Mowry, 46th IEEE/ACM International Symposium on Microarchitecture (MICRO-46), December 2013.
- “Communication-Efficient Distributed Multiple Reference Pattern Matching for M2M Systems,” Ruei-Bin Wang, Yu-Chen Lu, Mi-Yen Yeh, Shou-De Lin, and Phillip B. Gibbons, 13th IEEE International Conference on Data Mining (ICDM'13), December 2013.
- “Scalable, High Performance Ethernet Forwarding Lookup,” Dong Zhou, Bin Fan, Hyeontaek Lim, Michael Kaminsky, and David G. Andersen, 9th International Conference on emerging Networking EXperiments and Technologies (CoNEXT), December 2013.
- “Effective Mix-zone Anonymization for Mobile Travelers," Balaji Palanisamy and Ling Liu, Geoinformatica, Springer, online first, 2013.
- “Delay-tolerant Mix-zones on Road Networks,” Balaji Palanisamy, Ling Liu, Kisung Lee, Shicong Meng, and Yuzhe Tang, Distributed and Parallel Databases, Springer, online first, 2013.
LIST OF POSTERS
- “Recommending Memory Requests for Analytics," Charles Reiss and Randy Katz, Poster Session at the 4th ACM Symposium on Cloud Computing (SOCC’13), October 2013.
- “LiteMR: Fast Processing of Distributed Large Streaming Data,” Liting Hu and Karsten Schwan, Poster Session at the 4th ACM Symposium on Cloud Computing (SOCC’13), October 2013.
- “LazyTables: Distributed Machine Learning with the Stale Synchronous Parallel Model,” Henggang Cui, Qirong Ho, Jim Cipar, Jin Kyu Kim, Seunghak Lee, Wei Dai, Jinliang Wei, Greg Ganger, Phillip Gibbons, Garth Gibson, and Eric Xing, Poster Session at the 24th ACM Symposium on Operating Systems Principles (SOSP'13), November 2013.
- “Building a High-Performance Metadata Service by Reusing Scalable I/O Bandwidth,” Kai Ren, Swapnil Patil, Kartik Kulkarni, Adit Madan, and Garth Gibson, Poster session at the 24th ACM Symposium on Operating Systems Principles (SOSP'13), November 2013.
- “Agentless Cloud-wide Streaming of Guest File System Updates,” Wolfgang Richter, Canturk Isci, Benjamin Gilbert, Jan Harkes, Vasanth Bala, and Mahadev Satyanarayan, Poster session at the 24th ACM Symposium on Operating Systems Principles (SOSP'13), November 2013.
- “Scalable, High Performance Ethernet Forwarding Lookup on Commodity x86 Platform,” Dong Zhou, Bin Fan, Hyeontaek Lim, David G. Andersen, and Michael Kaminsky , Poster Session at the 24th ACM Symposium on Operating Systems Principles (SOSP), November 2013.
- “Hardware + Algorithms = Seriously Concurrent Hash Tables,” Xiaozhou Li, David G. Andersen, Michael Kaminsky, and Michael J. Freedman, Poster Session at the 24th ACM Symposium on Operating Systems Principles (SOSP), November 2013.
- “Tetrisched: Space-Time Scheduling for Heterogeneous Datacenters,” Alexey Tumanov, Timothy Zhu, Michael A. Kozuch, Mor Harchol-Balter, and Gregory R. Ganger, Poster Session at the 24th ACM Symposium on Operating Systems Principles (SOSP), November 2013.
List of presentations
In addition to the conference/workshop presentations associated with each of the published conference/workshop papers listed above, we had the following presentations:
- Ling Liu presented (GA Tech) presented a keynote at the IEEE 2013 Information Reuse and Integration (IRI) on ``Big Data Analytics as a Service: Exploring Reuse Opportunities,” San Francisco, CA, August 2013.
- Michael J. Freedman (Princeton) presented “Multi-tenant Resource Allocation for Shared Cloud Storage” at the New Results in Networking Research 2013, Microsoft Research, Redmond, WA, December 2013.
- Onur Mutlu (CMU) gave a keynote talk at Industry-Academia Partnership Stanford Cloud Workshop (http://www.industry-academia.org/event-stanford-workshop.html) on "Rethinking Memory System Design for Data-Intensive Computing," Mountain View, CA, December 2013.
- Phil Gibbons (Intel Labs) gave a distinguished lecture at EPFL on “The Intel Science and Technology Center for Cloud Computing,” Lausanne, Switzerland, December 2013.
ISTC Mission: Four inter-related research pillars (themes) architected to create a strong foundation for cloud computing of the future
The research agenda of the ISTC-CC is composed of the following four themes
- Specialization: Explores specialization as a primary means for order of magnitude improvements in efficiency (e.g., energy), including use of emerging technologies like non-volatile memory and specialized cores.
- Automation: Addresses cloud’s particular automation challenges, focusing on order of magnitude efficiency gains from smart resource allocation/scheduling and greatly improved problem diagnosis capabilities.
- Big Data: Addresses the critical need for cloud computing to extend beyond traditional big data usage (primarily, search) to efficiently and effectively support Big Data analytics, including the continuous ingest, integration, and exploitation of live data feeds (e.g., video or twitter).
- To the Edge: Explores new frameworks for edge/cloud cooperation that can efficiently and effectively exploit billions of context-aware clients and enable cloud-assisted client applications whose execution spans client devices, edge-local cloud resources, and core cloud resources.
Participants
Academic PI: Greg Ganger(CMU)
Executive Sponsor: Wen Hann Wang (CSR)
Managing Sponsor: Rich Uhlig (CSR-SAL)
Program Director: Jeff Parkhurst (APR)
Intel PI: Phil Gibbons
Intel Researchers:Michael Kiminsky, Mike Kozuch, Babu Pillai
Academic Partners: Dave Andersen, Guy Blelloch, Garth Gibson, Carlos Guestrin, Mor Harchol-Balter, Todd Mowry, Onur Mutlu, Priya Narasimhan, M. Satyanarayanan, and Dan Siewiorek (CMU); Mike Freedman, Kai Li, and Margaret Martonosi (Princeton); Anthony Joseph, Randy Katz, and Ion Stoica (UC Berkeley); Ada Gavrilovska, Ling Liu, Calton Pu, Karsten Schwan, and Sudha Yalamanchili (GA Tech).