Parallel Computing

Error with offloads from multiple threads causing crash

I have found an error when trying to perform offloads from multiple threads running the host, leading to a crash. I have created some test code to replicate this error. In this code many threads are created on the host, and each of these makes offload calls repeatidly, waiting for a hundred milli-seconds before making another. This does not cause a problem until the number of threads reaches around 140-150, at which point the program crashes with the following error (sometimes printed 2-3 times):

Does MIC have blend intrinsic?


I have two vectors, vector A: 1,3,5,7,9,11,13,15,17,19,21,23,25,27,29,31

vector B: 2,4,6,8,10,12,14,16,18,20,22,24,26,28,30,32

I want to generate two new vector:

new_a: 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16


Does MIC such intrinsic like blend such that I can achieve the above goal using two instructions in total?


how could I reuse memory allocated in sink side?


    I am using buffer to transfer data from source to sink. How could I resue the memory allocated in to sink side? I mean how could I implement the similar functionality as the following codes by using coi library:

   int a[1000];

   int b[100];

   #pragma offload tra

   for(int i = 0; i < 10; i++) {

   #pragma offload target(mic:0) in(a[i*100:(i+1)*100]: into(b:[0:100]) alloc_if(0) free_if(0))

   { ...}


   thanks a lot!

Can the Intel Phi communicate directly with a host NIC?

Hi -

With the Intel Phi, must all network traffic be brokered through the host? Or can one "assign" one of the NICs hosted in the PCI-Express bus and "give" it to the Intel Phi for it to use all by itself? Or must the host always act as a middleman for interrupts and packet delivery?

I don't (yet) have an Intel Phi to play with and my attempts at googling to produce a clear answer have failed. 



where should I put the library required by sink?


    After I run a sink program, I got the following error message:

    The remote process indicated that the following libraries could not be loaded:

     I guess I should upload to the mic. But where I should put this library? or how could I make this library uploaded automatically.

     Thanks a lot!

Phi fails after a certain number of offload+alloc calls

I have noticed what seems to be a bug when trying to run some code which makes a large(ish) number of offload calls. Too many offload (with alloc) calls over the entire program seems to be the problem, leading to a grinding halt (without error). I can replicate the problem with the code below, which just loops over a few thousand offload calls, allocating a very small amount of memory in each. The program stops dead after 1004 iterations, regardless of the upper limit on the loop, the size of the alloc, or any sleep time I put in the loop.

Subscribe to Parallel Computing