Subject: | |
From: | |
Reply To: | |
Date: | Wed, 15 Feb 2017 17:54:20 +0100 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Dear Spikey users,
The upgrade is completed and the Spikey system is available again.
Please find below some useful information on:
(1) Updated commands and basic workflow
(2) Latest version of the spikey module
(3) Developer tool for chip parameter manipulation
We hope that the new software and host computers work flawlessly for
you, as well. Please, contact us in case of problems or other questions.
With best regards,
Your Spikey Team
=== (1) Updated commands and basic workflow ===
Here is a line by line summary of the new work flow:
* Login at Heidelberg (new port number --> new host name is 'hel')
$ ssh [USER]@gitviz.kip.uni-heidelberg.de -p 11022
* Load software (now three commands; might change in the future)
$ .
/wang/environment/software/jessie/spack/current/share/spack/setup-env.sh
$ spack load --dependencies visionary-defaults
$ module load spikey
* Execute experiments (no changes)
$ cd spikey_demo/networks
$ srun -p spikey --gres station[NUMBER] python example.py
Note: Starting the python interpreter is currently very slow. We are
working on a solution and apologize for the inconvenience.
=== (2) Latest version of the spikey module ===
The command
$ module load spikey
will load the new default module. This module resembles the github
repositories dated 2017-02-08 and includes several bug fixes as well as
a new tool which is described below.
=== (3) Developer tool for chip parameter manipulation ===
In recent months, some experienced users have delved into low-level chip
configuration in more detail. For these users the following tool might
be of interest. We have developed a python module that permits direct
manipulation of low-level chip parameters from your pyNN script,
bypassing the pyNN/python software stack. The basic idea is to interrupt
pynn.run() after mapping the network to hardware resources, but before
triggering experiment execution on chip. Then manipulate parameters
directly and re-enter the pynn.run() routine.
In ipython, load the tweak module via
import pyNN.hardware.spikey.tools.tweak as tweak
Then type
tweak?
for details on usage in your pyNN script. As this is a developer module,
please be aware that it comes without in-depth support or "warranty".
########################################################################
To unsubscribe from the KIP-SPIKEY-USERS list, click the following link:
https://listserv.uni-heidelberg.de/cgi-bin/wa?SUBED1=KIP-SPIKEY-USERS
|
|
|