Calculation of MIPS
I have to calculate MIPS taken for my code. To calculate MIPS, I have a
formula ==> MIPS = (Total num of cycles/Total num of
If the formula is right, then, with the help of profiling in CCS, I can
find the cycles taken, but what about the instructions, should I do
Or, is there any way to find it out during profiling itself.
Waiting for your replies........
SASKEN BUSINESS DISCLAIMER
This message may contain confidential, proprietary or legally privileged information. In
case you are not the original intended Recipient of the message, you must not, directly or
indirectly, use, Disclose, distribute, print, or copy any part of this message and you are
requested to delete it and inform the sender. Any views expressed in this message are
those of the individual sender unless otherwise stated. Nothing contained in this message
shall be construed as an offer or acceptance of any offer by Sasken Communication
Technologies Limited ("Sasken") unless sent with that express intent and with due
authority of Sasken. Sasken has taken enough precautions to prevent the spread of
viruses. However the company accepts no liability for any damage caused by any virus
transmitted by this email
Does somebody has Codegen tools v4.32 for C6000 (I believe it is in CCS V2).
Could you send the build and link tools to me?
> I have to calculate MIPS taken for my code. To calculate MIPS, I have a formula
> MIPS = (Total num of cycles/Total num of Instructions)*CPU_Freq.
> If the formula is right, then, with the help of profiling in CCS, I can find the
> cycles taken, but what about the instructions, should I do manually??
> Or, is there any way to find it out during profiling itself.
Your formula wouldn't account for cases where some instructions are in parallel,
stalled due to external mem access, etc.
The way I usually do it is to relate processing time to input data. For example if
an input data frame arrives every 1 msec, and it takes the DSP 0.6 msec to process,
and TI rates the part at 4000 MIPS, then the algorithm is using 2400 MIPS. That
method is easy, and does offer the advantage of giving you a feel for the amount of
"overhead" remaining, but it still suffers from the problem of not really knowing how
many actual instructions are being executed.
If CCS could count the instructions, subtract out stalls, etc. then you could get a
more accurate figure.