English
资源说明：Spiking Neural Network implementation on CUDA platform
moment === Moment is a spiking neural network simulator that works on the [CUDA][cuda-link] platform. Rather than use pthreads or equivalents to parallelize the neurons, moment relies on the highly scalable and parallel cuda architecture. A question that might frequently arise is why choose to parallelize such an application? The answer lies in the final application of such a network. If a neural network is to be put to on-line use, the output should be generated as quickly as possible, as close to real-time as possible. Parallelization is indispensable for this kind of application. Let's see how this project proceeds. License --- Moment is licensed under the MIT License. Please have a look at the included LICENSE file. Using the test/visualizer program --- Moment includes a spike pattern viewer. It is written in Javascript and uses [jQuery][jq] and [Flot][flot]. First, either open the test/visualizer/index.html file directly in a browser or start a webserver by using something like ```shell python -m SimpleHTTPServer ``` while in the root directory. Once this server is running, navigate to `localhost:8000/test/visualizer` to see the interface. Copy _all_ the output of any of the test programs and paste it between the two square brackets in the interface. Hit draw to form a plot. There are 2 plotting areas provided so that you can compare two spiketrains. Using the build/format program --- The `build/format` program takes a file or a stream with one floating point number on each line, and converts that into a format that the Spiking Visualizer can easily use. ```shell tail -n+600 data/vib_table.data | head -n800 | ./build/format > a.js ``` Let's look at this part by part. The `tail` program will take the part of the file after the first 600 lines. This will be passed on to the `head` program, which will take the first 800 lines of the remaining lines. This 800 line unit will now be passed to the build/format program which will accordingly format it and write the output of the program to the a.js file. [cuda-link]: http://developer.nvidia.com/cuda [jq]: http://jquery.com/ [flot]: http://github.com/flot/
本源码包内暂不包含可直接显示的源代码文件，请下载源码包。
