Hello,
On TMS320F280x processors we often used the floating point fastRTS Library (sprca75) i.e. for fast sine calculations.Now we plan to use the AM335x processor. Is there a similar library?
How does the AM335x sine calculation benchmark compares to the TMS320C28346 (both devices running with full speed)?
ThanksBest regards,Patrick
Not officially that I know of, but we have taken some signal processing (starting with C) code (Matrix MPY and IIR) and run within Linux using the Neon co-processor as described here:
http://processors.wiki.ti.com/index.php/Cortex-A8#Neon_and_VFP_both_support_floating_point.2C_which_should_I_use.3F
Both came decently close to the theoretical best of 1 single-precision floating point mpy per 2 cycles as described here:
http://processors.wiki.ti.com/index.php/Feature_Comparison:_ARM_926,_1136_and_Cortex-A8
though (as expected) the Matrix MPY was much closer (since less data depenencies as with IIR).
And of course using Neon intrinsics worked better than pure C.
And the following DSP code was a nice reference:
http://processors.wiki.ti.com/index.php/C674x_DSPLIB