Guy Tel-Zur's Blog

 HPL Benchmark on my laptop It's the Top500 season time. I therefore tested HPL on my laptop using Intel's latest OneAPI version 2021.1.10.2261. The laptop specifications are obtained from lscpu : $ lscpu Architecture:                    x86_64 CPU op-mode(s):                  32-bit, 64-bit Byte Order:                      Little Endian Address sizes:                   39 bits physical, 48 bits virtual CPU(s):                          8 On-line CPU(s) list:     ...

The following no t e address es   questions regarding the  number of  required  NOPs in 5 stages MIPS pipelined processor.  In this example  lw  is the first instruction  follow ed  by an   add  (R-type) instruction with a RAW data dependency .     Table 1 In Table 1 we have an old MIPS which requires 3 NOPs because only after updating the architectural state (cycle 5) the add instruction can proceed.    (D)   refers to NOP instead of Decode.   Table 2 In Table 2 we assume that our  MIPS  can  write data in the first half of the clock cycle and read data in the second half  of the clock cycle  then the  number of NOPs  can be reduced  to 2.   Table 3 Finally in Table 3, like in the previous case, we assume the processor can  write data in the first half of the clock cycle and read data in the second half o...

Nice profiling using TAU from U. of Oregon This tool together with many other useful Parallel Processing tools and packages are available as a single VirtualBox ova image  here from Vi-HPS . You may also be interested to visit Paratools  (HPCLinux) website.

Parallel Debuggers are important tools when trying to eliminate bugs in parallel programs. Commercial parallel debuggers are expensive. In this blog post I will show you a free and open source alternative which is good enough for education purposes and for small codes. Suppose you want to debug the famous cpi.c code which computes an approximation to pi by a parallel numerical integration. This demo uses the free Data Display Debugger (ddd) which is based on gdb but has a nice GUI. Step 1, compile: mpicc -g -o cpi ./cpi.c Step 2, run (without a debugger): mpirun -np 2 ./cpi Step 3, run with the free Data Display Debugger (ddd) : mpirun -np 2 ddd ./cpi The last command will open two MPI processes (instances) of ddd each running cpi, see screen capture: Finally, I enclose here a short video which I hope is convincing: hi-res version:  https://youtu.be/N_J4NKJkuMs Happy (Free) MPI debugging!

OpenFOAM up and running in 5 minutes! Get the instructions and the container from  http://openfoam.com/download/install-binary.php   MYLAPTOP> ~/Downloads/startOpenFoam+ non-network local connections being added to access control list of_v1606_plus /opt/OpenFOAM bash-4.1$ echo 'now I am inside the container' now I am inside the container bash-4.1$ cd /home/telzur/OpenFOAM/telzur-v1606+/run/cavity/cavity blockMesh icoFoam paraFoam That's it!

sudo apt-get install subversion [sudo] password for telzur: Reading package lists... Done Building dependency tree       Reading state information... Done Some packages could not be installed. This may mean that you have requested an impossible situation or if you are using the unstable distribution that some required packages have not yet been created or been moved out of Incoming. The following information may help to resolve the situation: The following packages have unmet dependencies:  subversion : Depends: libsvn1 (= 1.8.8-1ubuntu3.2) but it is not going to be installed               Depends: libapr1 (>= 1.4.8-2~) but it is not installable               Depends: libaprutil1 (>= 1.2.7+dfsg) but it is not installable E: Unable to correct problems, you have held broken packages. Any ideas...?

Scientific Python versus Matlab: Geographical differences in interest  according to Google trends. Scientific Python: Matlab: Do you think what I think?

If you get error messages of the following kind: " E: Encountered a section with no Package: header. ..." Do the following: sudo rm /var/lib/apt/lists/* -vf sudo apt-get update It worked for me!

if you are in a situation where you want to install software using packagekit but nothing happens or you see a message that it is "waiting in queue" forever try the following which worked for me: Open terminal as root and type: pkcon refresh yum-complete-transaction this will let yum complete it's unfinished business and then packagekit will get back to life. Good luck

If you want to predict major breakthroughs in Supercomputing look at the sun. It is well accepted by now that every 11 years there is a 3 orders of magnitude mile stone in supercomputers. However, about 11 years is also the cycle of the Sun activity when counting for example the number of sunspots. Surprisingly enough whenever there is a deep in the sunspots count a major supercomputing milestone takes place. The chart below says it all The Wisdom behind Small Data Mining :-) Guy

HTCondor version 8.0 Windows Installation from gtelzur http://www.slideshare.net/gtelzur/condor8-win-install

You are welcome to join! https://www.facebook.com/groups/8450870046/

Forget about LINPACK and FLOP/s. It is the skin that matters. The next big thing should be the "Top500 Skin" ranking. Here are my top five:       Corollaries: 1. As can be seen the US maintains its dominance. 2. All vendors can try harder. How about consulting Italian car designers? 3. SC20 should take place at the Museum of Modern Art.

RaspberryPi , a $35 toy or the next disruptive technology and a possible building block for HPC ? This credit card sized board has a ARM11 processor and 512MB RAM. The board is powered via a micro USB connector. It has two built-in USB sockets, Ethernet and HDMI video. A schematic drawing of model B which I have can be found here . I installed  Raspbian  Linux, an optimized Debian version for Raspberry, on a 4GB SD card. A cool gadget for triggering thoughts about the future of computing. Enclosed below a few pictures I took, enjoy!

The Internet is becoming impossible ... Very sad

The slides of my talk at the  HPC Advisory Council Israel Supercomputing Conference 2012 that took place on February 7th, 2012 at Tel-Aviv university are available from here . The Importance of High-Performance and High-Throughput Computing in Israel

I tested the new HPC Europa Milestone 6 DVD and created a virtual cluster using VirtualBox on my laptop. I created a head-node and two computing nodes connected via a VLAN.  The head node has two network interfaces, NAT which enables the connection to the outside world and an internal virtual interface. The two computing nodes have only one network interface each for the VLAN. In picture number 1 the VirtualBox dashboard is shown with the 3 running computers Picture number 1: The VirtualBox dashboard showing the 3 running computers of the virtual cluster Picture number 2: The desktop of the head node (left) and the two computing nodes (right) In picture number two a PBS job executing a simple MPI task is demonstrated. The head node is on the left and the two computing nodes are on the right. My general impression is that this Virtual Cluster is an excellent tool for education where the teacher and later on the students can experience Parallel Processing using the...

I tested globus-provision which is an easy way to have a globus+condor cluster ready for number crunching on Amazon’s EC2 cloud. My front end was my laptop which runs Windows. Python on Windows does not behave exactly as on Linux when it comes to SIGINT and therefore os.fork() generated an error message. Thanks to the support of  Borja Sotomayor the SIGINT issues has resolved and the solution was to by-pass the SIGINT so I could proceed with the test. You can read more about this issue from here: http://jira.globus.org/browse/GP-13?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel I tried the example from: http://globus.org/provision/guide_compute_go.html#guide-compute-go I decided not to use globusonline for the eBooks transfer and I used wget instead. Below are a few screen captures that show the test. My configuration file:  Preparing the instance: after a minute or two: checking the instance: Watching my cluster at the AWS management ...

A year ago almost no one knew this term. With a doubling time of less than 6 months "Big Data" is becoming a hot hype, as can be seen below (by  Google Trends ) :

About four years ago I predicted in this blog the fall of Grid Computing and the rise of Cloud Computing (  http://telzur.blogspot.com/2007/10/end-of-grid-computing.html  ) Now I want to claim that Supercomputers as we know them today will not be the kind we will see in the future. I think that within a couple of years: 1. Most supercomputers will shift to the cloud (HPC Cloud). 2. The governance of X86 is not guaranteed: We already see the rise of GPGPUs and we will see more families of technologies being involved, e.g. Tegra and ARM, Reconfigureable Computing (RC) i.e. FPGAs, etc'. 3. Supercomputers won't be anymore General Purpose machines. They will be reconstructed as a Cloud Service ( Supercomputing as a Service ) from all the available technologies (see previous section) per single mission (program) in order to maximize the performance and minimize the consumed energy and cost. I decided to use the word mission because the word  task is u...