- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello!
I am trying to build a random-number generator, but I want to check the output values to see their distribution. I intend to write the random-numbers to the SRAM on board the DE2 board and view the contents using the Control Panel application. The catch is that I am using Linux. Is there a Linux distribution of the Control Panel application for DE2 boards? Is there any other way to view the on-board SDRAM/SRAM contents? Thank youLink Copied
4 Replies
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
As far as I know, there is no support for Linux with the pre-made GUIs. You can however use System Console to access your own custom logic, eg.,
http://www.alterawiki.com/wiki/using_the_usb-blaster_as_an_sopc/qsys_avalon-mm_master_tutorial Here's some random number generator documentation/code using LFSRs/PRBSs, http://www.ovro.caltech.edu/~dwh/correlator/pdf/lfsr_tutorial.pdf http://www.ovro.caltech.edu/~dwh/correlator/pdf/lfsr_tutorial_src.zip How are you generating random numbers? And for what purpose? Cheers, Dave- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you Dave!
I am from an ASIC design background, trying to cosy with FPGAs. I am trying to generate Gaussian samples starting with Tauseworthe Random Number Generators. I might dabble in hash functions later on too. I wanted to respond earlier, but the site seemed down - your response let me concentrate on other things without trying too hard to get Control Panel work on Linux using wine etc, and wasting a lot of time. Thank you, Anand- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
--- Quote Start --- I am trying to generate Gaussian samples starting with Tauseworthe Random Number Generators. --- Quote End --- Do you have references on this method? The few Google hits I looked at just showed LFSRs, which would not give you Gaussian noise, and if you've reviewed the link to the LFSRs tutorial, you'll see that you can get Gaussian-like noise, but its not really Gaussian (unless you sum enough uniform distros together so that the central-limit theorem works well). --- Quote Start --- I wanted to respond earlier, but the site seemed down - your response let me concentrate on other things without trying too hard to get Control Panel work on Linux using wine etc, and wasting a lot of time. --- Quote End --- Don't waste your time working with any of the closed-source applications provided with the evaluation boards. Cheers, Dave
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
This is a really late reply, sorry for that. I missed the e-mail from altera.
The reference I used was: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.97.9309&rep=rep1&type=pdf It takes the LFSR output and uses a Box-Muller transform on that involving sine,cos,log,sqrt etc to get a Gaussian Distribution. It looks like a very good candidate for FPGA implementation as you can get all these function implemented in a single cycle and throw in multiple parallel implementations to get a large number of samples per second. I did complete the coding and did some signal-taps using the Quartus software to sort of check what I got out of the design. The data looks roughly gaussian (didn't do any histogram checks - just judged that the samples had a large number of small values, and a small number of large values) - but I haven't had a chance to quantify it entirely, and a couple of things I saw tells me it's not quite right. The project is shelved for now, hoping to come back later to it. Yeah I stopped looking at terasic software. I just couldn't get anywhere with it. Maybe if I had gotten to your response earlier, I would have decided even earlier to not go on the wild goose cha Thanks Dave, Anand
Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page