Dear Pete, All, Just a small correction to what you say about EMIF and others, its document SPRU190D (TMS320C6000 Peripherals Reference Guide). Just wondering, is your statement true for the C64x family as well? BR, Indrajit --- pete <> wrote: > Hi, > > If you look at the SPRU109d under DMA and CPU data > access performance > you will see that CPU reads to external do not use > bursts, and hence > have this horrible access time. > > If you switch the cache on then you'll see a vast > improvement, but > beware the same issues occur with Data access from > external. Then you > have no cache on the C6201, so you need to use DMA > to move data onto and > off of the chip. > > Pete Warnes > Technical director > HUNT ENGINEERING (U.K.) Ltd > email > URL http://www.hunteng.co.uk or > http://www.hunt-dsp.com > phone +44 (0)1278 760188 > direct +44 (0)1278 760966 > GSM +44 (0)7711 190099 > Fax +44 (0)1278 760199 > > > -----Original Message----- > > From: goelhss <> > [mailto:] > > Sent: Friday, January 24, 2003 1:00 PM > > To: > > Subject: [c6x] C6201 - EMIF wait State ? > > > > > > Hi .... > > > > I have to run my code on 'C6201. > > If i use internal program memory, there is some > profiling > > figure, which is different than if i load my code > in external > > memory(i.e. SBSRAM, or SDRAM). > > > > Is it because of wait states ? > > The documents say that there is certain no. of > wait states > > for SBSRAM (1/2x or 1x cpu speed)or SDRAM(1/2x cpu > speed), > > depending upon settings like clock speed of > process (133 Mhz > > or 160Mhz.. in my case ). > > > > The difference of performance is huge.... it 's > taking nearly > > 25 times on ext mem, than in case of internal > memory. > > > > I thot of the culprit to be wait states.. and > tried to > > calulate the same, by singel stepping and using > > Profiler->Clock, which however, > > shows same load time for all memories... i.e. no > delay... but > > on the whole, entire code runs faster in internal > mem. > > > > Where am i goofing ? > > plz help... > > mail me @ . > > > > rgds, > > vishal. > > > > > > > > _____________________________________ > > Note: If you do a simple "reply" with your email > client, only > > the author of this message will receive your > answer. You > > need to do a "reply all" if you want your answer > to be > > distributed to the entire group. > > > > _____________________________________ > > About this discussion group: > > > > To Join: Send an email to > > > > > To Post: Send an email to > > > > To Leave: Send an email to > > > > > Archives: http://www.yahoogroups.com/group/c6x > > > > Other Groups: http://www.dsprelated.com > > > > > > ">http://docs.yahoo.com/info/terms/ > > _____________________________________ > Note: If you do a simple "reply" with your email > client, only the author of this message will receive > your answer. You need to do a "reply all" if you > want your answer to be distributed to the entire > group. > > _____________________________________ > About this discussion group: > > To Join: Send an email to > To Post: Send an email to > > To Leave: Send an email to > Archives: http://www.yahoogroups.com/group/c6x > > Other Groups: http://www.dsprelated.com > ">http://docs.yahoo.com/info/terms/ __________________________________________________

Hi, If you look at the SPRU109d under DMA and CPU data access performance you will see that CPU reads to external do not use bursts, and hence have this horrible access time. If you switch the cache on then you'll see a vast improvement, but beware the same issues occur with Data access from external. Then you have no cache on the C6201, so you need to use DMA to move data onto and off of the chip. Pete Warnes Technical director HUNT ENGINEERING (U.K.) Ltd email URL http://www.hunteng.co.uk or http://www.hunt-dsp.com phone +44 (0)1278 760188 direct +44 (0)1278 760966 GSM +44 (0)7711 190099 Fax +44 (0)1278 760199 > -----Original Message----- > From: goelhss <> [mailto:] > Sent: Friday, January 24, 2003 1:00 PM > To: > Subject: [c6x] C6201 - EMIF wait State ? > Hi .... > > I have to run my code on 'C6201. > If i use internal program memory, there is some profiling > figure, which is different than if i load my code in external > memory(i.e. SBSRAM, or SDRAM). > > Is it because of wait states ? > The documents say that there is certain no. of wait states > for SBSRAM (1/2x or 1x cpu speed)or SDRAM(1/2x cpu speed), > depending upon settings like clock speed of process (133 Mhz > or 160Mhz.. in my case ). > > The difference of performance is huge.... it 's taking nearly > 25 times on ext mem, than in case of internal memory. > > I thot of the culprit to be wait states.. and tried to > calulate the same, by singel stepping and using > Profiler->Clock, which however, > shows same load time for all memories... i.e. no delay... but > on the whole, entire code runs faster in internal mem. > > Where am i goofing ? > plz help... > mail me @ . > > rgds, > vishal. > > _____________________________________ > Note: If you do a simple "reply" with your email client, only > the author of this message will receive your answer. You > need to do a "reply all" if you want your answer to be > distributed to the entire group. > > _____________________________________ > About this discussion group: > > To Join: Send an email to > > To Post: Send an email to > > To Leave: Send an email to > > Archives: http://www.yahoogroups.com/group/c6x > > Other Groups: http://www.dsprelated.com > ">http://docs.yahoo.com/info/terms/

Hi .... I have to run my code on 'C6201. If i use internal program memory, there is some profiling figure, which is different than if i load my code in external memory(i.e. SBSRAM, or SDRAM). Is it because of wait states ? The documents say that there is certain no. of wait states for SBSRAM (1/2x or 1x cpu speed)or SDRAM(1/2x cpu speed), depending upon settings like clock speed of process (133 Mhz or 160Mhz.. in my case ). The difference of performance is huge.... it 's taking nearly 25 times on ext mem, than in case of internal memory. I thot of the culprit to be wait states.. and tried to calulate the same, by singel stepping and using Profiler->Clock, which however, shows same load time for all memories... i.e. no delay... but on the whole, entire code runs faster in internal mem. Where am i goofing ? plz help... mail me @ . rgds, vishal.