Rich is spending the week at ETH Zürich attending the Platform for Advanced Scientific Computing (PASC) Conference. Many thanks to two of the co-organizers, Markus Püschel and Olaf Shank, for being great hosts!
If you are interested in getting a copy of Rich’s talk slides + related material, see: hpcgarage.org/pasc14
The view from Villa Hatt at ETH Zürich is not too shabby!
Jee and Rich are in Phoenix this week for IPDPS’14, where Jee will present his latest updates to his “energy roofline” model [Tue May 20, Session 11] :
Rich also has the pleasure of serving in a panel discussion at the High-Performance, Power-Aware Computing workshop. His position is that power-aware scheduling is really an optimal control problem. For his 1-slide summary, see: [link]
Outside our group, Georgia Tech has a very strong representation this year. Looking at the program, we count 9 papers, including two of the four “Best Papers.” Go, Buzz!
Jee chowing down on a pre-conference pulpo at Mariscos Sinaloa.
Piyush, Jee, and Rich are in Portland representing the lab at the SIAM Parallel Processing ’14 meeting. If you are also there, please drop by our various events! List below; resources (papers, slides) are here: [link].
This week Rich has the great fortune of attending Dagstuhl Seminar 13401: Automatic Application Autotuning for HPC Architectures. He’ll be summarizing Jee’s work (with critical assists by Marat) on the energy archline model. For his slides and pointers to relevant materials, look here: [www]
“Classical” wing of Schloss Dagstuhl
We’re in Seattle this week attending the US DOE Workshop on Modeling and Simulation of Exascale Systems and Applications — “ModSim’13” [www]. A copy of Rich’s talk and pointers to supplemental materials is available here: hpcgarage.org/modsim13
Jee and Kent are speaking on Wednesday, May 22 at IPDPS’13. If you are attending, be sure to check out their talks!
Both sessions are in the William Dawes Room at the Hyatt Regency Cambridge. Hope to see you there!
Rich gave a keynote at the 3rd International Workshop on Accelerators and Hybrid Exascale Systems (AsHES) earlier this week. A copy of his slides is available here: hpcgarage.org/ashes2013
A notional space of processors.
Kent and Jee will each present a paper at IPDPS’13, to be held in Boston in May. Kent’s paper is about a unique take on high-level algorithm-architecture co-design; Jee’s paper is about his “roofline” energy model. If you plan to attend, be sure to see their talks! We’ve also posted preprints under Papers.
Aparna and Jee are at SIAM CSE 2013 in Boston this week, giving talks the fast multipole method and the energy roofline, respectively. (UPDATE: Click on the title for Jee’s talk to access the PDF slides.)
Rooflines in time vs. “archlines” in energy
Check out our new technical report, which presents a thought-experiment on the question of whether engineering an algorithm to optimize time differs from doing so with respect to energy (e.g., Joules).
Abstract
We describe an energy-based analogue of the time-based roofline model of Williams, Waterman, and Patterson (Comm. ACM, 2009). Our goal is to explain—in simple, analytic terms accessible to algorithm designers and performance tuners—how the time, energy, and power to execute an algorithm relate. The model considers an algorithm in terms of operations, concurrency, and memory traffic; and a machine in terms of the time and energy costs per operation or per word of communica- tion. We confirm the basic form of the model experimentally. From this model, we suggest under what conditions we ought to expect an algorithmic time-energy trade-off, and show how algorithm properties may help inform power management.
Subject area: High-Performance Computing
Keywords: performance analysis; power and energy modeling; computational intensity; machine balance; roofline model
Note (January 14, 2013): The above link to the PDF now points to the official version on the Georgia Tech Library’s report repository. If you downloaded an earlier version (dated December 24, 2012), you may wish to re-download this newer version. (The only difference is the addition of Appendix B, made on December 31, 2012.)
Jee is bringing The HPC Garage one step closer toward being an actual garage! Here he is taking some fine-grained power measurements using PowerMon 2.0, a cool device for taking DC power measurements. PowerMon 2.0 is the work of some of the fine folks at RENCI.
Jee is PowerMan! (with PowerMon)
This work is part of our study to determine the precise relationships between computational time and energy. (Thanks, Jee, for putting up with my insistence that we document this event. — Rich)