Threading

best way to compute products of the form A*S*A' where S is symmetric?

What's the best way to efficiently compute a product of the form A*S*A' where S is symmetric with MKL?

Does MKL provide anything similar to the sykrx function that's part of cublas (http://docs.nvidia.com/cuda/cublas/#cublasxt_syrkx)? so that you can use it along with symm to compute the product without having to recompute symmetric entries?

How to prevent multi-rate filter from scaling output

I'm using a lowpass filter created with ippsFIRMRInit to upsample a signal by a user definable factor. The output of the filter is scaled roughly (not exactly) proportional to the upsample factor.

Is this expected behavior? If so, how do you determine the correction factor to make the upsampled data look like the original signal?

There's a "doNormal" option for ippsFIRGenLowpass, but there's nothing in the documentation that explains what it does. I tried setting it to both true and false, with no noticeable difference in the output.

scif_open: Permission denied

Hi,

System: Ubuntu 14.04.1 LTS, MPSS 3.4.1, parallel_studio_xe_2015_update1 and  g++ (Ubuntu 4.8.2-19ubuntu1) 4.8.2.

I am trying to use scif for communication with mic. However on the host node, I got the following error when calling scif_open

Error no: 13, msg: Permission denied

However when I run problem in sudo mode, it works fine.

How can I fix this problem?

  Thanks

Severe (157): Program Exception - access violation

When I try to run any of the 3 analyses available in Advisor I get the error "Severe (157): Program Exception - access violation" and the line with the call of "omp_get_num_procs" is listed in the console output:

use omp_lib
integer*4 procs
procs = omp_get_num_procs ( )

I double checked that I use following settings in my project: /debug:full, /Od, /check:none, /libs:dll /threads /dbglibs (Debug Multithread DLLs), /Qopenmp.

Power for multiple Xeon Phi cards

Hi,

System:  Ubuntu 14.01 LTS, MPSS 3.4.1

We have a server with 2KW power supply, which the manufacture claims support up to 6 Xeon Phi Cards.

When I run my program on Xeon Phi (31S1P) uses all the cores (100% core utilization). I see the power usage on each card is about 188  Watts, which is much lower than 275W in the Xeon Phi specification sheet. I wonder if I can plug all 6 Xeon Phi 31S1P card to the server. What should I check to make sure the system is not under power?  In case the system is under power, is there any damage to CPU and GPU?

   Thanks

over-heating issues

I have a 31S1p which over-heats. I put a 70 CFM fan in the front of it, and built a duct, as was suggested in this forum. I put a 40 CFM fan on the back, and the card still over-heats. I even tried opening the chassis, and blasting a floor fan directly at the card, and it over-heats. Is it possible that the card is defective? I don't know what else to try here.

Phi Supported Motherboards

Hi,

I've just bought a Phi as part of the promotion and am looking to do some code optimization and energy measurement on it. I need to build a workstation and have been looking at appropriate motherboards. I understand that MMIO above 4G (or above 4G decoding) is required in the BIOS. Is this all that is required? Is there a list of motherboards that will support the Phi?

I've seen that the ASUS Sabretooth Z97 Mark 2 supports this option in the bios has anyone had any experience of using this board to build a small workstation? Any other suggestions welcome.

Bug in icc 2015.1.148???

My code:

#include <stdio.h>

int dd, dm, dy, i;
char s[32], d[5];
FILE *file;

int main() {

  file = fopen("d.txt", "r");

  for(i=0; i<5; i++) {

    fgets(s, 30, file);

    d[0]=s[0];
    d[1]=s[1];
    d[2]=0;
    dd=atoi(d);

    d[0]=s[3];
    d[1]=s[4];
    d[2]=0;
    dm=atoi(d);

    d[0]=s[6];
    d[1]=s[7];
    d[2]=s[8];
    d[3]=s[9];
    d[4]=0;
    dy=atoi(d);

    printf("%i %i %i %s", dd, dm, dy, s);

Subscribe to Threading