Many clouds use COTS ethernet, eg. AWS, Alibaba, Oracle. My expectation
is that most workloads that are tightly coupled are on hundreds of
nodes. Networking is disaggreagated, so variance in latency will be
somewhat greater than a typical cluster, though some of the newer
topologies used in HPC c
On 6/28/23 11:18, Tony Travis wrote:
> On 28/06/2023 07:18, John Hearns wrote:
>> Rugged individuaiist? I like that... Me puts on plaid shirt and
>> goes to wrestle with some bears,,,
>>
>> > Maybe it is time for an HPC Linux distro, this is where
>> Good move. I would say a lightweight distro
On 3/19/22 1:28 PM, Mikhail Kuzminsky wrote:
Just in the HPCG discussion, it was proposed to use the now widely
used likwid benchmark to estimate memory bandwidth. It gives
excellent estimates of hardware capabilities.
Am I right that likwid uses its own optimized assembler code for each
specifi
For memory bandwidth, single node tests such as Likwid are helpful
https://github.com/RRZE-HPC/likwid
MPI communication benchmarks are a good complement to this.
Full applications do more than the above, but these are easier starting
points that require less domain specific application knowled
On 4/23/21 7:07 AM, Greg Lindahl wrote:
I'm buying a 100 gig ethernet switch for my lab, and it seems that the
latest gear is intended to run a switch OS. Being as cheap as I've
always been, free software sounds good.
It looks like Open Network Linux is kaput.
It looks like SONiC is doing pret
GNU Parallel ( http://www.gnu.org/software/parallel/ ) might allow for
similar workflows
On 11/21/20 3:56 AM, Lux, Jim (US 7140) via Beowulf wrote:
If Joe has interpreted your need correctly, I’ll second the suggestion
of pdsh – it’s simple, it works pretty well, it’s “transport”
independent (
But even then, it was a pretty slow evolution – the Fortran compilers I
was running in the 80s on microcomputers under MS-DOS wasn’t materially
different from the Fortran I was running in 1978 on a Z80, which wasn’t
significantly different from the Fortran I ran on mainframes (IBM 360,
CDC
On 10/13/20 3:12 PM, Oddo Da wrote:
Jim, Peter: by things have not changed in the tooling I meant that it is
the same approach/paradigm as it was when I was in HPC back in the late
1990s/early 2000s. Even if you look at books about OpenMPI, you can go
on their mailing list and ask what books t
that
will definitely be handy.
--
Prentice
On 8/7/20 5:02 AM, Benson Muite wrote:
Maybe the following are helpful:
https://sx-aurora.github.io/posts/hpcg-tuning/
https://www.hpcadvisorycouncil.com/pdf/HPCG_Analysis_POWER8.pdf
https://link.springer.com/chapter/10.1007/978-3-030-50743-5_21
htt
Maybe the following are helpful:
https://sx-aurora.github.io/posts/hpcg-tuning/
https://www.hpcadvisorycouncil.com/pdf/HPCG_Analysis_POWER8.pdf
https://link.springer.com/chapter/10.1007/978-3-030-50743-5_21
https://ulhpc-tutorials.readthedocs.io/en/latest/parallel/hybrid/HPCG/
https://www.intel.c
Dear Doug,
Might you be willing to give some indication of benchmarks you find useful for
customers of Limulus systems?
Regards,
Benson
On Sat, Feb 22, 2020, at 6:42 AM, Douglas Eadline wrote:
>
> That is the idea behind the Limulus systems -- a personal (or group) small
> turn-key cluster t
Enabling local manufacturing using CAD and additive manufacturing may be
helpful.
Machine learning applications such as voice recognition, image recognition and
recommendation systems might also be interesting. For example one could improve
library catalog search.
On Sat, Feb 22, 2020, at 1
YES! Though national infrastructures such as XSEDE may be easier to sustain and
for community colleges to use.
On Wed, Feb 19, 2020, at 10:46 PM, Mark Kosmowski wrote:
> Is there a role for a modest HPC cluster at the community college?
> ___
> Beowul
On Tue, Feb 11, 2020, at 9:31 AM, Skylar Thompson wrote:
> On Sun, Feb 09, 2020 at 10:46:05PM -0800, Chris Samuel wrote:
> > On 9/2/20 10:36 pm, Benson Muite wrote:
> >
> > > Take a look at the bootable cluster CD here:
> > > http://www.littlefe.net/
> &g
On Mon, Feb 10, 2020, at 9:46 AM, Chris Samuel wrote:
> On 9/2/20 10:36 pm, Benson Muite wrote:
>
> > Take a look at the bootable cluster CD here:
> > http://www.littlefe.net/
>
> From what I can see BCCD hasn't been updated for just over 5 years, and
> the la
Take a look at the bootable cluster CD here:
http://www.littlefe.net/
On Mon, Feb 10, 2020, at 1:54 AM, Mark Kosmowski wrote:
> I purchased a Cisco UCS C460 M2 (4 @ 10 core Xeons, 128 GB total RAM) for
> $115 in my local area. If I used ESXi (free license), I am limited to 8 vcpu
> per VM. Coul
Charges depend alot with field. People in medical fields generally pay $2,000
per paper for open access, with their research grants supporting this - the
idea being that medical information is useful and it is cheaper for the funding
body to make their research widely available, than for librari
Generally, getting published does not depend on having an academic
qualification if the work is sufficiently interesting. Choose venue
appropriately. There seems to be quite a plit between domain specific
publications and HPC publications, with only a few venues able to reliably
review both the
a) For technically knowledgable audience, you could demonstrate some
simple benchmarks codes.
b) For a more general audience, you might also look for some parallel
open source applications that are specific to the domain of interest.
For example for engineering, OpenFOAM (https://openfoam.com/
MPI is not likely the best programming model at present for such a
scenario, because fault tolerance is not great. However, there are
people running many loosely coupled parallel programs on clusters
because the computational capacity and storage mechanisms make it
convenient, even if the inter
There are some machine learning workloads that could be done in this
context. Renting smartphones may be more challenging, but desktop or
server idle computing power could be used in distributed offline
training algorithms.
The physics community does make heavy use of grid computing
(https://
There are some machine learning workloads that could be done in this
context. Renting smartphones may be more challenging, but desktop or
server idle computing power could be used in distributed offline
training algorithms.
The physics community does make heavy use of grid computing
(https://
Hi,
You might check work by Ryohei Kobayashi:
https://dblp.org/pers/hd/k/Kobayashi:Ryohei
Benson
___
Beowulf mailing list, Beowulf@beowulf.org sponsored by Penguin Computing
To change your subscription (digest mode or unsubscribe) visit
https://beowu
Hi Faraz,
Mellanox manuals can be found at:
https://docs.mellanox.com/
Example setup instructions (not sure if correct for you as do not have
exact details on your hardware):
https://www.mellanox.com/related-docs/prod_software/Mellanox_OFED_Linux_User_Manual_v4_3.pdf
Maybe also helpful (stu
Hi Faraz,
Have you tried any other MPI distributions (eg. MPICH, MVAPICH)?
Regards,
Benson
On 4/30/19 11:20 PM, Gus Correa wrote:
It may be using IPoIB (TCP/IP over IB), not verbs/rdma.
You can force it to use openib (verbs, rdma) with (vader is for
in-node shared memory):
mpirun --mca b
"Many employers look for people who studied humanities and learned IT by
themselves, for their wider appreciation of human values."
Mark Burgess
https://www.usenix.org/sites/default/files/jesa_0201_issue.pdf
On 2/23/19 4:30 PM, Will Dennis wrote:
Hi folks,
I thought I’d give a brief introdu
Welcome! Do you also contribute to documentation - even improving grammar would
be helpful?
On 2/23/19 4:30 PM, Will Dennis wrote:
Hi folks,
I thought I’d give a brief introduction, and see if this list is a good fit for
my questions that I have about my HPC-“ish” infrastructure...
I am a ~30y
There are a number of tools. A possible starting point is:
http://spcl.inf.ethz.ch/Research/Scalable_Networking/SlimFly/
Regards,
Benson
On 1/4/19 12:44 AM, Alexandre Ferreira Ramos wrote:
Hi Eveybody,
Happy new year!
I need to conduct a study on simulating the performance of a large scale
c
Thoughts and papers related to computer benchmarks are sought for the
Benchmarking in the data center workshop. More details at:
http://parallel.computer/
___
Beowulf mailing list, Beowulf@beowulf.org sponsored by Penguin Computing
To change your subscr
Some power transformer oils do work - but performance measurements,
performance/price evaluation and long term system durability studies
seem to be lacking.
On 11/5/18 10:16 AM, Tony Brian Albers wrote:
> Salt water is highly corrosive, that's why people use mineral or
> silicone oil.
>
> I've hea
On 10/11/18 10:08 PM, Douglas Eadline wrote:
> All:
>
> Over the last several months I have been reading about:
>
> 1) Spectre/meltdown
> 2) Intel Fab issues
> 3) Supermicro MB issues
>
> I started thinking, if I were going to specify a
> single rack cluster, what would I use?
>
> I'm assuming a g
Thanks. This is interesting. Possibly also of interest:
https://clearlinux.org/
Not for HPC, but some aspects look useful.
On 09/10/2018 06:50 AM, John Hearns via Beowulf wrote:
Chris Samuels recent post reminds me.
I went to a fascinating and well delivered talk by Jason Hearne McGuiness
https
On 06/19/2018 09:47 PM, Prentice Bisbal wrote:
On 06/13/2018 10:32 PM, Joe Landman wrote:
I'm curious about your next gen plans, given Phi's roadmap.
On 6/13/18 9:17 PM, Stu Midgley wrote:
low level HPC means... lots of things. BUT we are a huge Xeon Phi
shop and need low-level programme
Hi Faraz,
Last looked at OpenFOAM 4-X
https://github.com/OpenFOAM/OpenFOAM-4.x
Though development has migrated to
https://develop.openfoam.com/Development/OpenFOAM-plus
Steps I used in a bash build script for 4-X on a cluster with OpenMPI,
qt, zlib and cmake already installed are below. Scaling
Consider trying:
https://github.com/amd/blis
https://github.com/clMathLibraries/clBLAS
as well.
On 02/23/2018 12:48 AM, Prentice Bisbal wrote:
Just rebuilt OpenBLAS 0.2.20 locally on the test system with GCC 6.1.0,
and I'm only getting 91 GFLOPS. I'm pretty sure OpenBLAS performance
should be
There is a very nice and simple Max flops code that requires much less
tuning than Linpack. It is described in pg 57 of:
Rahman "Intel® Xeon Phi™ Coprocessor Architecture and Tools"
https://link.springer.com/book/10.1007%2F978-1-4302-5927-5
An example Fortran code is here:
https://github.com/bk
May also be of interest:
JobDigest – Detailed System Monitoring-Based Supercomputer Application
Behavior Analysis
Dmitry Nikitenko, Alexander Antonov, Pavel Shvets, Sergey Sobolev,
Konstantin Stefanov, Vadim Voevodin, Vladimir Voevodin and Sergey Zhumatiy
http://russianscdays.org/files/pdf1
On 08/17/2017 09:54 PM, mathog wrote:
On 17-Aug-2017 11:10, Alex Chekholko wrote:
The Google paper from a few years ago showed essentially no correlations
between the things you ask about and failure rates. So... do whatever is
most convenient for you.
This one?
http://research.google.c
This was nice (support has ended, but still runs):
http://open-mx.gforge.inria.fr/
Comparisons with other things would be interesting.
Benson
On 03/17/2017 04:33 PM, Lux, Jim (337C) wrote:
This is a bit of a blast from the past question.. Back when Beowulfs
were the “new hot thing”, we all u
Hi,
Do you have any performance requirements?
Benson
On 02/13/2017 09:55 AM, Tony Brian Albers wrote:
Hi guys,
So, we're running a small(as in a small number of nodes(10), not
storage(170TB)) hadoop cluster here. Right now we're on IBM Spectrum
Scale(GPFS) which works fine and has POSIX suppo
40 matches
Mail list logo
