Intel® VTune™ Amplifier XE

2017 Beta feature: ring buffer collection

Ever wanted to collection data for the last x seconds/minutes of your application's execution so that you didn't have tons of data to sift through?!  Well, now you can do exactly that with the VTune Amplifier XE 2017 beta software.

From the command line, use the ring-buffer option.  From the graphical user interface, use the Limit collected data by option in the Advanced section of the Analysis Target tab.

Calling Python Developers - High performance Python powered by Intel MKL is here!

We are introducing a Technical Preview of Intel® Distribution of Python*, with packages such as NumPy* and SciPy* accelerated using Intel MKL. Python developers can now enjoy much improved performance of many mathematical and linear algebra functions, with up to ~100x speedup in some cases, comparing to the vanilla Python distributions. The technical preview is available for everybody at no cost and includes early access to profiling support in VTune Amplifier.

LLC miss count and QPI traffic

Hi Folks,

Recently I'm involved in a project to investigate the NUMA awareness and CPU affinity of a new file system. My method is to choose a baseline file system that is well-known to be NUMA friendly, and then compare the QPI traffic with the new file system by writing the same amount of data.

I have a question about the QPI traffic and the miss count of last level cache fulfilled by remote DRAM and cache. I thought they must be strong correlated as QPI traffic should be caused by remote memory access, until I saw the following stats from vtune:

Track the memory usage as a function of time

Hi forum users,

I have an executable and I would like to track its memory usage as a function of time. I tried the "memory access" view point, but could not extract this information.

The only relevant post I found on this topic is an ancient one.

Is this possible with VTune? What is the correct analysis set up? Any comment is appreciated.


Line numbers of call stack in command line version

I am working on a linux server developing an application and profiling it using Vtune (2016). Within the network of my organization it is feasible to use the GUI version over ssh. This is often not possible on the train or airplain connection and there I have to use the command line version. This would normally not be a problem, besides that:

project navigator default ON

Is there a way to open with only Welcome window when vtune opens for the first time? I don't find the project navigator tree on LHS to be very useful and would like it to not show up if possible. I tried View--> options on RHS options menu but they don't persist if I close vtune and re-open it.

I can't run hardware event-based sampling analysis

Hello. I've been tried running enable-call-counts knob by specifying the code on pbs script. 

amplxe-cl -collect-with runsa -knob enable-call-count=true -data-limit=0 knob ./raxmlHPC -m GTRCAT -n 23S.E.ALL -s 23S.E/R0/cleaned.alignment.fasta -j -p 6154307

When I run vtune. I keep getting error saying it cannot load data collected.

What am I doing wrong with my code?

订阅 Intel® VTune™ Amplifier XE