Roadrunner Supercomputer to be Decomissioned

  • Thread starter Thread starter CFDFEAGURU
  • Start date Start date
Click For Summary
SUMMARY

The Roadrunner supercomputer, known for its groundbreaking performance, is set to be decommissioned, marking the end of an era in high-performance computing. Users in the discussion express interest in supercomputing capabilities, with mentions of building custom systems using dual Intel Xeon E5 processors and 64GB of RAM. The conversation highlights the importance of multi-threading and parallel processing, as well as the relevance of CUDA for Nvidia and OpenCL for AMD in software development. Resources like the Parallella project are suggested for those interested in accessible supercomputing solutions.

PREREQUISITES
  • Understanding of multi-threading and parallel processing concepts
  • Familiarity with Intel Xeon E5 architecture
  • Knowledge of CUDA and OpenCL programming frameworks
  • Basic awareness of supercomputing technologies and their applications
NEXT STEPS
  • Research building custom high-performance computing systems with dual Intel Xeon E5 processors
  • Explore the Parallella project for affordable supercomputing options
  • Learn about CUDA programming for Nvidia GPUs
  • Investigate OpenCL for AMD hardware compatibility
USEFUL FOR

This discussion is beneficial for hardware enthusiasts, computational fluid dynamics analysts, and developers interested in high-performance computing and supercomputing technologies.

CFDFEAGURU
Messages
781
Reaction score
10
I have always wanted to work with a supercomputer. My CFD (computational fluid dynamics) analyst has 128 paralleled processors at his office and that is as close as I have come so far.

Maybe I could buy this one? :)

http://phys.org/news/2013-03-line-roadrunner-supercomputer.html

Thanks
Matt
 
Computer science news on Phys.org
there are some people doing some weird things with ARM chips.
http://www.kickstarter.com/projects/adapteva/parallella-a-supercomputer-for-everyone?ref=category

multi-threading, multi-cores, parallel, those concepts can be confusing. you could either make it multi-seats while more people sharing the same computer or one person using more than one computer.

I think I would just build my own computer with dual xeon e5 with 64gb of ram, add one of these if it's not enough:
http://www.intel.com/content/www/us/en/high-performance-computing/high-performance-xeon-phi-coprocessor-brief.html

if the software it's related to cuda, then Nvidia , opencl then AMD.
 

Similar threads

  • · Replies 4 ·
Replies
4
Views
7K
Replies
11
Views
4K
Replies
10
Views
4K
  • · Replies 21 ·
Replies
21
Views
3K
  • · Replies 17 ·
Replies
17
Views
5K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 14 ·
Replies
14
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 5 ·
Replies
5
Views
3K