Starting out, with 2 to 12 Phi's

So I bought 2 of the 31S1P's (not yet in use) and am contemplating getting another 10 for a fluid dynamics simulation. I'm trying to figure out how to proceed. I apologize if some of these are stupid questions. I think they're reasonable though as my setup is a bit unique: luckily I have a few 3d printers, in case I need to print brackets or ducts, and a 3600 CFM fan with 15" diameter. Some questions:

Help with camera preview in Android media sample

Hello Intel INDE team,

I am building an app with live streaming and am using code from the Android sample CameraStreamerActivity. Everything works great in landscape mode and I get a full screen preview. However, when I try to run this in portrait mode, I get a square preview which does not fill the screen. I've tried setting different preview sizes and even removing setPreviewSize, but I still cannot fill the whole screen.

This is a sample of how my preview looks: http://imgur.com/XUznvF1


Thank you.



problems in offload in fortran modules

Hi all,

I have been trying to run what is called CESM(climate earth system model) and added an offload to one of its Fortran modules.

!dir$ offload begin target (mic)
    print *,'hi this is the offload section'
!dir$ end offload

I performed a successful build. The ar I replaced with the xiar and managed a successful build of the modules.

However while running the final executable i get the following error.

Allow for C++11 move semantics in TBB mutex and lock types

Please make tbb::queuing_mutex and tbb::queuing_rw_mutex non-movable in C++11 and later to meet C++11 requirements for STL mutexes (§30.4.1). Mutex implementations outside the STL are not required to meet the requirements for the STL, but this is probably what the developers would expect.

In addition, if possible, please also implement move constructors and move assignment operators for the respective scoped_lock inner classes. The std::lock_guard and std::unique_lock classes have this functionality. If this is not possible, please delete these methods.


MPSS 3.4.3 restart failed when try to add an NFS mount entry to mic0


the host running on centos 7 has a new kernel Linux 3.10.0-123.20.1.el7.x86_64,

before mpss installation, I recompiled the mpss module and reflashed the firmware to 391 for mpss 3.4.3,

mic and mpss run well by using miccheck, mininfo, systemctl status mpss. 

NFS server has been installed in the host with:  yum install nfs-utils

rpcbind and nsf service are running well. firewalld and NetworkManger have been stopped. SELinux is disabled.

Thread Error Analysis: Intel Inspector and boost::thread::once.hpp corrupts stack


I tried to use Intel Inspector (from XE 2015) with little success. The application I try to analyze uses boost threads and a boost spirit parser. Thread safety defines have been added (eg BOOST_SPIRIT_THREADSAFE).

The following example displays the problem. In normal Release/Debug builds the program will output :

$ build64/Debug/once_test.exe


If Intel Inspector XE 2015 is used with "Thread Error Analysis" the program fails with stack corruption during the execution of boost::call_once.

infiniband connection host-mic and mic-mic

Hi, I'm trying to set up infiniband connection between host and mic, mic and mic. Host is showing this on ifconfig:

mic0:ib: flags=67<UP,BROADCAST,RUNNING>  mtu 64512
        inet  netmask  broadcast
        ether 4c:79:ba:20:06:63  txqueuelen 1000  (Ethernet)

and two: mic0 and mic1 interfaces.

I have two coprocessors installed.  I can run ib_read_bw between host and mic0, but not host and mic1 or mic0 and mic1.  Getting error:  

Received 10 times ADDR_ERROR
Unable to perform rdma_client function.

Subscribe to Optimization