|
Cutting complexity
|
20/06/2008
|
| |
Digital signal processing is an interesting term and means different things to different people. An additional complication is that dsp is beginning to disappear as the technology is absorbed into various applications.
Dr Johannes Stahl is vice president of marketing and business development for electronic system level (esl) design software specialist CoWare. He said there has always been confusion over what dsp means. “On the one hand, there are off the shelf parts, such as those from Texas Instruments. On the other hand, digital signal processing is a generic term and has nothing to do with a programmable processor.
“There’s a range of things we look at in the design of dsp systems. They could be hardware, they could be software but what drives them all are applications that require more bandwidth and higher performance.”
In Dr Stahl’s opinion, there was once a definite difference between dsps and dsp architectures. “But the fields are merging and designers don’t always use classical dsp architectures any longer – even companies like TI see an end of life for that approach.”
What this means, he believes, is a move to new architectures. “For example, very long instruction word (vliw) and parallel architectures will be used going forward to execute dsp algorithms.”
And it’s because of increasing algorithmic complexity that CoWare has announced a strategic relationship with Agility Design Solutions that is focused on accelerating the simulation and implementation of complex dsp algorithms – particularly those starting from Matlab models. Agility’s mission, in a nutshell, is to reduce the time needed to develop, implement and verify signal processing algorithms.
|
| |
Author
Graham Pitcher
|
| |
| |
|
This material is protected by Findlay Publications copyright 2009.
See Terms and Conditions.
One-off usage is permitted but bulk copying is not.
For multiple copies contact the sales team.
|
| |
|
| Email this article |
| |
|
|
|
|