DOE ASCR workshop.
Friday, 29 July 2011 15:01
administrator
Rich is attending the DOE-ASCR Workshop on Exascale Programming Challenges, a discussion-focused pow-wow among programming model luminaries, focused on sorting out what may (or may not?) need to change in order to help build and maintain scalable applications for the machines expected to appear in the 2018-2020 timeframe.
 (View from Mulholland Dr., over-looking the Hollywood Bowl, the 101, and downtown L.A.)
CRUISE Interns
Wednesday, 20 July 2011 17:05
Richard Vuduc
The lab thanks Lesly and Gurbinder for joining us over the summer, as CRUISE program interns. We learned a lot from the studies they carried out and wish them the best of luck in all their future endeavors!
ISC'11, live!
Tuesday, 21 June 2011 08:05
administrator
The lab is in Hamburg this week for ISC'11. Watch our tumblr log (hpcgarage.tumblr.com) for updates. UPDATE: Rich's talk slides on "Balance principles for algorithm-architecture co-design" are available here (PDF).
HotPar'11.
Monday, 30 May 2011 16:19
Richard Vuduc
Casey, Kent, and Aparna are back from HotPar'11, where they presented our work on balance principles for "co-design." Update: Slides posted (see below). Enjoy!
- K. Czechowski, C. Battaglino, C. McClanahan, A. Chandramowlishwaran, R. Vuduc. Balance principles for algorithm-architecture co-design. In Proc. USENIX Wkshp. Hot Topics in Parallelism (HotPar), Berkeley, CA, USA, May 2011. [PDF | PDF slides]
DiGPUFFT ("dig-puffed") - Distributed GPU 3D FFT
Saturday, 14 May 2011 14:38
Richard Vuduc
Chris has released DiGPUFFT (pronounced, "dig-puffed"), a set of code patches that add GPU capabilities to P3DFFT, Dmitry Pekurovsky's distributed memory 3D FFT library. A technical report describing these patches and our performance results for the Keeneland system at ORNL is in the works; it should appear on the DiGPUFFT website later this month or early next.
Talk at Berkeley.
Sunday, 24 April 2011 14:19
administrator
Rich just finished a short trip to Berkeley, where he had a chance to talk about some of the work that the HPC Garage did on the GT-led Gordon Bell winner from last year.
The highlights included meeting not just the "giants" of computing research, but a short excursion to those of nearby Muir Woods:
|
ICIAM'11
Saturday, 23 July 2011 12:42
administrator
Rich is back from the 2011 SIAM International Congress on Industrial and Applied Mathematics (ICIAM) [twitter]. Never before have so many applied math nerds been in one place! Rich spoke in the four-part Minisymposium on Creating the Next-Generation of High-Performance Numerical Computing Capabilities (Parts I, II, III, and IV), which featured a diverse peek into a future of sophisticated, composable, scalable, and automatically generated and tuned code for scientific simulation. For Rich's talk slides on Concurrent Collections (CnC) for HPC, click here: [PDF]
CAV (EC)^2
Friday, 15 July 2011 13:14
Richard Vuduc
Just got back from the (EC)2 workshop in Snowbird, co-located with the Computer Aided Verification conference. This was a great opportunity to learn about bridging topics between performance-oriented HPC and formal verification. You'll find our talk slides here: [PDF]
AFDS'11.
Tuesday, 21 June 2011 08:02
administrator
The HPC Garage is back from Seattle / Bellevue, where we participated in the AMD Fusion Developer Summit. Rich gave a panel talk, which previewed some of the lab's on-going work on algorithm-architecture co-design (PDF slides). Also, the HPC Garage band, "Casey and the Bloodhounds," made its debut. YouTube video forthcoming, so stay (performance) tuned!
CnC'11: Join in the fun!
Monday, 16 May 2011 12:46
Richard Vuduc
The 3rd annual Concurrent Collections (CnC) workshop will take place on Sep. 7, 2011 in Fort Collins, Colorado, USA, co-located with LCPC'11. This meeting is a forum for researchers and developers of parallel programs to interaction on a variety of issues related to next-generation parallel programming models, particularly those that use dataflow-programming ideas. Please visit the workshop website (cnc11.hpcgarage.org) to find out how you can participate. Abstracts (~ 200-500 words) are due by July 31.
Salishan 2011.
Sunday, 01 May 2011 19:51
administrator
 Rich is back from the US Dept. of Energy-sponsored 2011 Salishan Conference on High-Speed Computing, where he had a chance to preview some of the thought-provoking exascale forecasting work being done by Kent, Casey, and Chris. (Work currently under review; stay tuned!)
HotPar'11 bound.
Saturday, 05 March 2011 00:37
Richard Vuduc
We'll be presenting a paper at this year's HotPar workshop, on the topic of "co-design":
- K. Czechowski, C. Battaglino, C. McClanahan, A. Chandramowlishwaran, R. Vuduc. "Balance principles for algorithm-architecture co-design." In Proc. USENIX Workshop on Hot Topics in Parallelism (HotPar), May 2011. (accepted)
|
