Michael-
> best to switch to the c5509 /07/03 dsp. These work
as expected,
I wish we could. But we need both HPI and EMIF active at the same time.
> I can tell ...
> I was told the purpose of the c5501/02 was to work for mobile communcation
,
> so the data extern will be transfered to intern memory, where these
devices
> run really with 300 MHz.
I think we can get our C5502 board to work Ok. The application is for real-time
IIR
parametric filter bank, so the amount of code that is critical for real-time
performance is fairly small, and can fit into i-cache. We can keep filter
coefficients in onchip memory. So probably it will run Ok.
But yes I agree, for other applications, C5501/2 is not a good choice.
> BTW: Using code composer - if you have a hardware
at your dsp - doesnot make
> sense (to express it reputably). Not even the CSL lib is realy cool.
Only
> thing, which makes things easier, is the Makro - support of the CSL. We had
a
> latency for an external interrupt, using code composer (c5503, 200 MHz)
> + dsp-bios 1.6s
> + standard CSL disppatcher 600 ns
> + register programming 60 ns
Yes I know about these latencies, but I can tell you other DSPs have similar
issues.
CCS + JAG is simply not that fast. People make too many assumptions about the
debug
environment and what CCS can do. We tend to rely on HPI for "quick"
debug and fast
data transfer, especially when DSP code is running.
-Jeff
EMIF performance
Started by ●November 18, 2005
Reply by ●January 31, 20062006-01-31
