Audio Weaver is the drag-and-drop graphical design software.
“DSP Concepts has ported their platform software, AWE Core [embedded processing engine], on to xcore.ai platform that runs along with Xmos platform software,” Xmos told Electronics Weekly. “This is transparent to the user, except they have to have this binary running on Xmos development kit. The user then downloads Audio Weaver platform on windows, builds a DSP audio pipeline from a large stock of existing modules, and runs the pipeline real-time on the xcore.ai MultiChannel audio development board.”
Xcore-ai ICs have two 800MHz hardware cores, each with 512kbyte ram and a vector unit, on which are implemented 16 logical cores that share up to 1,200Mips, or 2,400Mips in dual issue mode.
The vector units are capable of up to eight whole-word, 16 half-word or 32 byte multiply-adds, as well as quad complex multiply or 256 bit-wide multiply-adds.
This architecture is suited to both DSP and AI processing, according to Xmos.
“Audio innovation should be easy so that the audio engineer can focus on the differentiation for the products they are making,” said DSP Concepts CEO Chin Beckmann. “Audio Weaver is purpose-built to streamline the design process. A visual approach, multiple deployment options, and integrations with chip partners all help to keep the focus on the product, and not on a frustrating development process. Xmos’ adaptability only adds to that strength.”
XK-AUDIO-316-MC-AB USB MultiChannal Audio evaluation kit is the development hardware mentioned above.