MIPS usage
Thread Summary
The user inquires about the impact of MIPS usage on temperature, power consumption, and lifespan of the ADAU1467 processors. The engineer explains that while temperature can affect lifespan, the difference between 69% and 93% MIPS usage is not significant. Balancing MIPS usage is not critical unless there are concerns about program completion within sample periods, especially with external clocking and looping code.AI Generated Content
Hello,
I have a big project that use 2 ADAU1467
The compiler gives me MIPS usage for both, one is at 69% one is at 93%
So I have a few questions regarding MIPS usage :
- I guess that the chip's temperature and power consumption are related to the usage ? I couldn't find any clue in the datasheet.
- Also I couldn't find any info regarding life span of the chips, am I right thinking the MIPS usage has an impact of the expected life span, or not at all ?
The underlying question here is : is there any advantage of trying to balance the MIPS usage in both, or none at all ?
I've found no documentation regarding the protduct life span, or temperature, or any advice regarding MIPS consumption.
If I've missed some documentation please refer me to it.
Thanks !
Audio Signal Processors
Datasheet/Specs
adau1467
processors and dsp
SigmaDSP Audio Processors
Processors and Microcontrollers
Show More
Hello JBAX,
Good questions but it is not mentioned in the datasheet because it is not significant.
Yes, temperature will have some small effect on lifespan but that would be really tough to measure and get specs. Unless we are talking temperatures that are really high, higher than the absolute min and max.
You do not want to risk one of the processors from not finishing the program on a sample period that is exceptional. Umm... This is a little bit complicated to explain. It also depends on your program. It depends on how you are clocking the core? IF you set the start pulse to an external rate. AND use your local master clock from a crystal, then then have a lot of looping code then there is a chance that you might have fewer instructions on some sample periods. More on some. What this core does if there are not enough clock cycles is it will just stop the program and start back to the beginning of the program. This means some of the outputs might not get any update on that sample period. Which means the output sample will be repeated. Tough to hear. We have an overrun monitor which can capture this but I think if you are using internal start pulse for the sample rate then 93% should be fine.
The core will not run that much hotter. It will run a little hotter but not a lot. 69% and 93% is not all that different. It is not like 10% and 93%. So if your solution is done and working then leave it alone. If you want to try to balance it out a little go ahead but it is not critical or important to do so.
Seach the forum for Overrun Monitor and you should find some posts about it. I am out of town at a trade show so I do not have time.
By the way, this sounds like a really interesting project!!
Thanks,
Dave T
Please log in to reply.