LEA Data Memory
(Shared with system)
16 bit to 32 bit Bridge
Multilevel Harvard Engine
LEA Core
Peripheral Interface
Memory Interface
LEA Internal Memory
- Commands
- Coefficients
- Constants
MSP430 Peripheral Bus
MSP430 Memory Bus
LEA Module
32 bit
16 bit
LEA Introduction
447
SLAU367P – October 2012 – Revised April 2020
Copyright © 2012–2020, Texas Instruments Incorporated
Low-Energy Accelerator (LEA) for Signal Processing
17.1 LEA Introduction
The LEA is a 32-bit hardware engine designed for operations that involve vector-based signal processing,
such as FIR, IIR, and FFT without CPU intervention. The LEA supports multiple commands, which are
issued by CPU. The LEA offers incomparable performance and energy consumption when performing
vector-based digital signal processing computations; for performance benchmarks comparing LEA to using
the CPU or other processors, see
Benchmarking the Signal Processing Capabilities of the Low-Energy
.
Figure 17-1. LEA System Block Diagram
17.2 LEA Operation
The LEA begins executing the selected operation when the CPU writes a LEA command to the LEA
command register when the LEA is in idle mode. Before writing the command, the CPU must configure the
LEA argument registers with the pointers to the parameter blocks for the designated operation. The LEA
performs the operation without CPU intervention and triggers an interrupt when the operation is complete.
The LEA accesses the LEA data memory, which is used for input data, output data, and the parameter
blocks. The LEA data memory is also accessible by the CPU and the DMA, so that the output data of the
LEA operation can be moved to other memory location by the CPU or the DMA (see
). The
CPU and the LEA can run simultaneously and independently unless they access the same system
memory (RAM). See the device-specific data sheet for details about LEA availability and LEA data
memory size.