tag:blogger.com,1999:blog-15011337.post3975834993449208028..comments2023-01-10T13:28:58.576-05:00Comments on Bob McGwier N4HY: OMAP processor/ Beagle BoardBob N4HYhttp://www.blogger.com/profile/00592369079956478688noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-15011337.post-58675554437090629172009-02-16T07:11:00.000-05:002009-02-16T07:11:00.000-05:00Bob: My thesis work, back in the early 1980's, in...Bob:<BR/> My thesis work, back in the early 1980's, involved mapping speech processing algorithms on an SIMD machine. Back then I would have never guessed that today people would own SIMD machines and not even know it!<BR/><BR/>What's a good source of information on the NEON architecture? I've looked at the ARM site, but mostly found press releases.<BR/><BR/>The NEON might make a good case study for my architecture class.<BR/><BR/>--Mark, ka9elgAnonymoushttps://www.blogger.com/profile/02754964756036714335noreply@blogger.comtag:blogger.com,1999:blog-15011337.post-91059871336791005622009-02-13T09:01:00.000-05:002009-02-13T09:01:00.000-05:00Hello Mark!The answer is yes or no depending on wh...Hello Mark!<BR/><BR/>The answer is yes or no depending on what you mean. Several of us have the Beagle Boards and we are running Ångström on the boards. But the real work is not on DttSP per se. The real work is doing the NEON SIMD coding of several blocks. That work IS being done but not directly to DttSP. It is being done directly for GnuRadio by Phil Balister and I am doing some other NEON work as well. Once this NEON math work is done, DttSP will just compile and run.Bob N4HYhttps://www.blogger.com/profile/00592369079956478688noreply@blogger.comtag:blogger.com,1999:blog-15011337.post-34268977762690832282009-02-13T07:32:00.000-05:002009-02-13T07:32:00.000-05:00Bob: Porting DttSP to the BeagleBoard sounds inte...Bob:<BR/> Porting DttSP to the BeagleBoard sounds interesting. Has anything been started?<BR/><BR/>--Mark, ka9elgAnonymoushttps://www.blogger.com/profile/02754964756036714335noreply@blogger.com