Computational Science Stack Exchange
Computational Science Stack Exchange

Questions tagged [hpc]

High Performance Computing generally uses large arrays of computers connected in clusters or grids.

HPC systems usually come in 3 varieties:

  1. Supercomputers.
  2. Cluster computers, sometimes called Beowulf clusters are generally built out of a number of regular computers in order to get the throughput of supercomputers.
  3. Grid computers, are also generally built out of a number of regular computers.

The main difference between Grids and Clusters have to do with the communication between computing nodes (a node will be a CPU core). If there will be a lot of communication, one generally calls it a cluster. In contrast, Grids tend to have little communication between nodes.

146 questions
17
votes
4 answers

Profiling CFD code with Callgrind

I'm using Valgrind + Callgrind to profile a solver I have written. As the Valgrind user manual states, I've compiled my code with the debugging options for the compiler: "Without debugging info, the best Valgrind tools will be able to do is guess…
tmaric
  • 1,916
  • 1
  • 11
  • 22
15
votes
4 answers

Should I rent computing resources, or buy my own computers

Since this question is related to computation, I decided to post here. Hopefully it will be seen as appropriate. I've just started running atmospheric and oceanic models, and I realize that I need more cores, memory, and disk space than my current…
user4624937
  • 265
  • 2
  • 5
7
votes
1 answer

Online resources for reviewing graphics cards for GPGPU

Can anyone recommend a site that maintains up-to-date reviews of graphics cards for GPGPU use? Most benchmarks focus on gaming performance, whereas I am interested in the performance of scientific applications like FFTs, linear algebra, and…
Emre
  • 333
  • 1
  • 7
7
votes
5 answers

External hardware resources for running long and computationally intensive simulations

I've written code for an obstructed random walker simulation and I want to run long simulations (6 hours or more on my computer). I don't want to run this simulation on my computer because I will want to run multiple simulations simultaneously and I…
Ami
  • 241
  • 1
  • 4
4
votes
1 answer

Apply for a cluster for scientific computing from a developing country?

I don't have access to a computer cluster in my university. Is there website that accepts applications for free access to a computer cluster for scientific computing? Further information: I am in Cuba. I need something on the order of 30 processors,…
a06e
  • 1,729
  • 15
  • 22
3
votes
3 answers

Can anyone recommend a pipeline dependency ("make") tool for SGE jobs?

I am looking for a tool to specify pipelines of arbitrary multiple jobs with dependencies for the Sun Grid Engine. I can wire things together myself but I am hoping for something more declarative and reusable. Can anyone recommend such a tool or…
Alex Stoddard
  • 131
  • 3
3
votes
1 answer

Pricing of HPC resources in terms of power usage

Following the latest trends in linear algebra, power consumption is quickly becoming an important factor to consider when designing and using an algorithm. I've even heard that over time, the electricity bill can exceed the initial purchasing cost…
Thijs Steel
  • 1,693
  • 8
  • 16
2
votes
2 answers

HPC: Intel QPI versus AMD with HyperTransport

I am looking into upgrading to a High Performance Computing work-station to do mostly computational fluid dynamics computations and some fundamental numerics. The CFD software that is available does not support GPU/CUDA. Aside from sufficient RAM…
seb
  • 994
  • 1
  • 9
  • 22
2
votes
1 answer

Get job script directory with SGE

With the Sun Grid Engine, how can I get the directory where the job script is located, within the job script? I need this so I can start the script with cd script-location to make sure it'll have access to all the necessary files. So far I have…
Szabolcs
  • 2,620
  • 2
  • 19
  • 34
2
votes
0 answers

Collaborative coding platform in the style of Google Colab

(I hope this question is fitting for this SE, I couldn't find very relevant tags) I am looking for an online platform in the style of Google Colab with the following requirements: possibility to run notebooks like in JupyterLab necessity of at…
bela83
  • 443
  • 1
  • 3
  • 12
2
votes
1 answer

What is the minimum raw FLOPS that supercomputers need to be included in the Green500 list?

The Green500 List lists the top 500 most energy efficient supercomputers. My question is, what is the minimum FLOPS that computers need to be considered for the Green500 list? Do they just take the TOP500 list and rank them according to FLOPS/watt…
Ciro Santilli OurBigBook.com
  • 121
  • 4
1
vote
0 answers

Will Sumatra interfere with git?

As far as I understand, Sumatra does some sort of versioning of simulation scripts and output files. If I start to use Sumatra inside a git working directory, can Sumatra end up interfering with git or vice versa?
Thomas Arildsen
  • 321
  • 2
  • 6
1
vote
3 answers

Flush output in Torque scheduler?

I'm trying the Torque scheduler. I have previous experience with the LSF scheduler, and I quite like being able to use 'bsub -I make -j 12' to compile a program really quickly, i.e. running an 'interactive' job, and seeing any errors as they…
Hugh Perkins
  • 373
  • 1
  • 13
1
vote
1 answer

Where can an undergraduate go to find cores on a budget?

I've may have reached a point in my neural network research that I cannot continue without significant financial investment. I am using neuroevolution to evolve a network on the EMNIST data set. It is embarrassingly parallel (100 individuals *…
user27075
1
vote
1 answer

Requesting less than a node with SLURM

I am curious if it is possible to request less than a node using SLURM. For example, if I use the commands, #!/bin/bash #SBATCH --job-name="name" #SBATCH --output="name.%j.%N.out" #SBATCH --ntasks 1 #SBATCH --time=00:30:00 will SLURM automatically…
EssentialAnonymity
  • 599
  • 1
  • 3
  • 16
1
2