Differences

This shows you the differences between two versions of the page.

--- technical:whitepaper:start [2014-09-24 10:34] – external edit 127.0.0.1
+++ technical:whitepaper:start [2022-06-17 12:24] – frey
@@ Line 1: / Line 1: @@
-====== White Papers ======
-{{ :technical:whitepaper:1344435822_27-edit_text.png?128|}}Some of the content in this area will be in PDF format and may need to be downloaded before being read.
-------
-===== Mills: threading performance study =====
-The behavior of the Mills cluster's cutting-edge Interlagos processor is studied under multi-threaded and multi-process work loads.  Influences of compiler and BLAS/LAPACK library choice are presented.
-{{:technical:whitepaper:mills-threading-performance-study.pdf|Download the PDF}}
-===== Mills: AMD Opteron 6200 Unix Tuning Guide =====
-The Nodes on the Mills cluster have 2 or 4 AMD Opteron 6200 series sockets.  Each socket is organized as a multi-chip module package with two CPU dies, interconnected using a HyperTransport link.  Each die is organized as 3 core pairs (Interlagos modules). Thus, to the OS, the socket appears as a 12 logical CPUs (12-core sockets). Resources such as memory and floating points unites are shared between the cores.
-This technical tuning guide is intended for "systems admins, application end-users,
-and developers on a Linux platform who perform application development, code tuning, optimization, and initial
-system installation".  The document describes resource sharing, and the effect on your applications.
-[[http://amd-dev.wpengine.netdna-cdn.com/wordpress/media/2012/10/51803A_OpteronLinuxTuningGuide_SCREEN.pdf|Download the PDF from the AMD developer site]]
-===== HPC Challenge Awards Competition at SC Conference =====
-The SC((The International Conference for High Performance Computing, Networking, Storage and Analysis)) High Performance Computing Challenge includes the benchmarks:
-  - HPL  measures the floating point rate of execution for solving a linear system of equations
-  - DGEMM  measures the floating point rate of execution of double precision real matrix matrix multiplication
-  - STREAM  measures sustainable memory bandwidth (in GB/s) and the corresponding computation rate for a simple vector kernel
-  - PTRANS (parallel matrix transpose)  exercises communications between pairs of processors. It is a useful test of the total communications capacity of the network.
-  - Random Access  measures the rate of integer random updates of memory (GUPS)
-  - FFT  measures the floating point rate of execution of double precision complex one dimensional Discrete Fourier Transform (DFT)
-  - Communication bandwidth and latency  measures latency and bandwidth of a number of simultaneous communication patterns; based on b_eff (effective bandwidth benchmark).
-[[http://hpcchallenge.org|Download the PDF]]
-===== Matlab: Computational threads on a shared cluster =====
-By default Matlab uses multiple computational threads for standard linear algebra calculations.  Without the options ''-singleCompThread'' it will use libraries tuned to use the computational hardware.  Examples are the ''sunperf'' library on Solaris (Strauss) and the MKL library on intel hardware including Mills.
-To fully use the computational threads you must call the built in high level functions or data parallel constructs in Matlab.  For example, it is easy to write loops to do a Matrix multiply, but it w
-===== Mills: Using ACML In High Performance Computing Challenge =====
-For Mills, the recommended libraries include OpenMPI, ACML, and FFTW.  The AMD recommended compilers include Open64 and PGI.
-The following document from AMD includes instructions for installing these libraries, but this is not needed on Mills since they are already installed as VALET packages.
-[[http://developer.amd.com/wordpress/media/2012/10/ACMLinHPCC.pdf| Download the PDF from the AMD developer site]]
-===== Mills: Benchmarking studies =====
-==== High Performance Computing Challenge studies ====
-  * [[hpcc open64 acml|open64 compiler with ACML and openmpi libraries]]