potatotron
Well-known member
I'm going to play with bitbanding to see if it's faster than ANDing and ORing bits in bytes.
Bit banded RAM is only available in the upper block (SRAM_U) -- is there an easy way to force the compiler to allocate an array at a specific RAM address?
In case you're interested in details, I have 64x32 arrays of HSV data I need to convert to temporal frames for APA102s at 60fps. I've optimized the conversion down to about 50 milliseconds but it still needs to be about 4x faster to reach 60fps.
Bit banded RAM is only available in the upper block (SRAM_U) -- is there an easy way to force the compiler to allocate an array at a specific RAM address?
In case you're interested in details, I have 64x32 arrays of HSV data I need to convert to temporal frames for APA102s at 60fps. I've optimized the conversion down to about 50 milliseconds but it still needs to be about 4x faster to reach 60fps.