http://172.24.10.4/gpu-workshop-2012/ 2026-05-02T16:31:06+02:00 http://172.24.10.4/gpu-workshop-2012/ http://172.24.10.4/gpu-workshop-2012/lib/tpl/default/images/favicon.ico text/html 2012-03-08T16:44:18+02:00 http://172.24.10.4/gpu-workshop-2012/directions?rev=1331221458&do=diff The venue is the LMU Biocenter which is located at the HighTech Campus Martinsried. Its address is: LMU Biocenter Großhaderner Str. 2 82152 Planegg-Martinsried Germany It is best reached by public transport (see below). Map Google Map of the surrounding: text/html 2012-03-08T14:49:09+02:00 http://172.24.10.4/gpu-workshop-2012/hotels?rev=1331214549&do=diff Hotel Thalmair Heiglhofstr. 3, 81377 München Tel. 089 / 741163-0, Fax 089 / 741163-41 <http://www.hotel-thalmair.de> Prices: Double room: 110 € / night, Single room: 88 € / night, Breakfast is included. Location: Close to the public transport: stop “Großhadern” (U6, Bus 266) text/html 2012-04-23T12:57:26+02:00 http://172.24.10.4/gpu-workshop-2012/pictures?rev=1335178646&do=diff text/html 2012-04-23T12:59:30+02:00 http://172.24.10.4/gpu-workshop-2012/start?rev=1335178770&do=diff News * April 23, 2012: Pictures from the symposium. * April 12, 2012: The workshop topics are online. * April 9, 2012: Eugene Izhikevich from our sponsor Brain Corporation will announce job opportunities at the symposium. * April 2, 2012: The program for the developer workshop is online. * March 19, 2012: The preliminary program is online. * March 8, 2012: We now have directions and hotel suggestions for you. * March 7, 2012: The list of symposium speakers is fina… text/html 2012-04-13T12:25:31+02:00 http://172.24.10.4/gpu-workshop-2012/symposiumprogram?rev=1334312731&do=diff Symposium program Time Speaker Title 09:00 - 09:05 Thomas Wachtler Welcome 09:05 - 09:10 Michael Schmuker Introduction 09:10 - 09:50 Romain Brette Models and challenges for neural simulation on GPU 09:50 - 10:30 Giovanni Idili An OpenCL-based Hodgkin-Huxley solver for the OpenWorm multi-algorithm simulation engine 10:30 - 10:50 -- Coffee break -- 10:50 - 11:30 David Higgins Using OpenCL to simulate a large network of Leaky Integrate-and-Fire neurons with plastic synapses … text/html 2012-04-13T14:06:42+02:00 http://172.24.10.4/gpu-workshop-2012/topicbenchmarking?rev=1334318802&do=diff Description The Tesla cards support memory error correction (ECC), which is on by default. ECC is known to come with a significant cost on performance. Regarding the surprisingly good performance of the NVidia GTX 580 in Andeas' benchmarks, we hypothesized that the reason might be that the Tesla card Andreas used had ECC turned on, and that it would be significantly faster with ECC turned off, probably faster than the GTX580. text/html 2012-04-13T14:06:16+02:00 http://172.24.10.4/gpu-workshop-2012/topicbrian2genn?rev=1334318776&do=diff Thomas Nowotny and Dan Goodman (not to be blamed for eventual errors in this synopsis as it was written by Michael Schmuker) Goals Provide Brian to a high-performance GPU-enabled backend using GeNN. Brian has already such a backend in form of the NeMo backend, but GeNN complements NeMo's capabilities for modular network architectures and dense connections. In addition, GeNN supports neuronal models which NeMo doesn't. text/html 2012-04-12T15:23:34+02:00 http://172.24.10.4/gpu-workshop-2012/topicbrian2nemo?rev=1334237014&do=diff Andreas Fidjeland et al. Goals Results Outlook Back to workshop topics text/html 2012-04-12T15:20:53+02:00 http://172.24.10.4/gpu-workshop-2012/topiccableequations?rev=1334236853&do=diff Remy degenne et al. Goals Results Outlook Back to workshop topics text/html 2012-04-13T16:10:01+02:00 http://172.24.10.4/gpu-workshop-2012/topiccodegeneration?rev=1334326201&do=diff Dan Goodman et al. Goals Results Outlook Resources The documentation for code generation in the SVN version of Brian is available here: <http://dl.dropbox.com/u/556426/brian-dev-docs/docs/developer-codegen.html> Back to workshop topics text/html 2012-04-13T10:26:20+02:00 http://172.24.10.4/gpu-workshop-2012/topicgenn2pynn?rev=1334305580&do=diff Michael Schmuker, Thomas Nowotny, Pierre Yger Goals Provide a PyNN frontend for GeNN. Progress Thomas is implementing a Python frontend for GeNN in order to make GeNN interact with Brian . We're waiting for the Python frontend to be useable, since it will also be the basis for the PyNN backend. text/html 2012-04-11T22:15:32+02:00 http://172.24.10.4/gpu-workshop-2012/workshop?rev=1334175332&do=diff Insert content here