MSP430 Peripherals
Hardware Accelerators
One question that is often asked, "Why would you put dedicated hardware accelerators onto low-
cost, low-power processors?"
It's an interesting question ... with a very practical answer. If a specific functionality is required,
accelerators are the most efficient implementation. Take for example, the CRC or AES modules;
serial (and wireless) communications are often requiring these functions to make the data
transmissions robust and secure. To implement these functions in software is possible, but would
actually consume a lot more power. Further, the memory footprint for an algorithm (code and
data) often ends up greater than the smaller footprint of the hardwired accelerator. Thus, where it
makes sense, you'll see TI adding dedicated hardware modules.
Another example is the multiplier. We can benefit from it without any programming effort, since
the compiler automatically uses this hardware, when it's available.
With regards to the Direct Memory Access (DMA) peripheral, we caution you ... if you find
yourself using memcpy() in your code, you should investigate how the DMA might save you time
and power. It also should be utilized in your peripheral driver software whenever and wherever it's
available.
MSP430 Workshop - Introduction to MSP430 1 - 27
To Next Page
Loading...
Need help?
General
