Case study: optimizing PPA with RISC-V custom extensions in TWS earbuds
By John Min, Andes Technology
embedded.com (August 11, 2022)
A look at using custom extensions in a RISC-V processor to enable power, performance and area (PPA) optimized true wireless stereo earbuds SoC design.
A common goal for many IC designs is achieving an optimum combination of power, performance, and area (PPA). This article examines the components and design of a currently shipping true wireless stereo (TWS) chip design, indicating the engineering tradeoffs made to achieve the PPA demanded by the final ear buds. In particular, we will also look at the RISC-V ISA (instruction set architecture) extensions that helped to dramatically reduce power while achieving the desired performance at no extra cost.
A TWS chip is mostly a Bluetooth RF transceiver. It comprises a 128-tap finite impulse response (FIR) filter, 128-tap least means square (LMS) adaptive filter, an active noise control (ANC)
hybrid function and the computational elements and buses needed to compute these functions. In this design, the first three of the four elements would require the greatest amount of computing resource: The FIR filter will demand 1,693 instruction cycles, the LMS filter another 1,820 cycles, and the ANC function, a whopping 10,404 cycles. The system block diagram illustrates the major components of the TWS design with analog functions on the left and digital on the right. The Bluetooth transceiver (top left) consists of an RF baseband controller and 64kbytes of exchange memory. The transceiver and audio subsystem below are complex because the TWS design has two independently operating speakers and microphones. The phase lock loop (PLL) for each must be synchronized to ensure each is running at the same frequency.
To read the full article, click here
Related Semiconductor IP
- Multi-core capable 64-bit RISC-V CPU with vector extensions
- 64 bit RISC-V Multicore Processor with 2048-bit VLEN and AMM
- RISC-V AI Acceleration Platform - Scalable, standards-aligned soft chiplet IP
- 32 bit RISC-V Multicore Processor with 256-bit VLEN and AMM
- All-In-One RISC-V NPU
Related White Papers
- Creating a custom processor with RISC-V
- Extending RISC-V ISA With a Custom Instruction Set Extension
- A guide to accelerating applications with just-right RISC-V custom instructions
- ISA optimizations for hardware and software harmony: Custom instructions and RISC-V extensions
Latest White Papers
- CRADLE: Conversational RTL Design Space Exploration with LLM-based Multi-Agent Systems
- On the Thermal Vulnerability of 3D-Stacked High-Bandwidth Memory Architectures
- OmniSim: Simulating Hardware with C Speed and RTL Accuracy for High-Level Synthesis Designs
- Balancing Power and Performance With Task Dependencies in Multi-Core Systems
- LLM Inference with Codebook-based Q4X Quantization using the Llama.cpp Framework on RISC-V Vector CPUs