will this DSP class help me find intern/job more easily?

--------------HAPPY NEW YEAR TO ALL DSPERS ---------

Hi all,

I am considering taking a class in DSP with emphasis on TI processors.

The course description states: "This course introduces the fundamentals of 
applied digital signal processing by implementing a wide range of digital 
signal processing (DSP) applications on the Texas Instrument (TI) TMS320C54x 
DSP processor."

It will surely improve my understanding of DSP in general. But it also 
involves the workload of TMS DSP programming, although my interest is more 
on algorithm design...

What do you think? Can DSPer in this forum give me some experiences? Will 
this help me find an summer intern/job more easily?

Thanks a lot guys! 

kiki wrote:
> --------------HAPPY NEW YEAR TO ALL DSPERS ---------

Thanks and likewise.

> Hi all,
>
> I am considering taking a class in DSP with emphasis on TI
processors.
>
> The course description states: "This course introduces the
fundamentals of
> applied digital signal processing by implementing a wide range of
digital
> signal processing (DSP) applications on the Texas Instrument (TI)
TMS320C54x
> DSP processor."
>
> It will surely improve my understanding of DSP in general. But it
also
> involves the workload of TMS DSP programming, although my interest is
more
> on algorithm design...
>
> What do you think? Can DSPer in this forum give me some experiences?
Will
> this help me find an summer intern/job more easily?

Well, I don't know what other experience you have with computer
programming, so I'll try to be careful not to be too definitive in
my advice. I don't have much experience with DSP programming as such,
but I did learn a lot of assembly programming on the then state-of-
the-art i8086 - 80286 + x87 maths processors in the late 80ies/early
90ies.

I've hardly done assembly programming after taking those classes,
but the experience and insights are invaluable. I know a lot more
about what happens inside the computer, what to do and what not to do,
what makes a program fast and what slows it down. These days I don't
do much manual optimization to speed things up, but I do my best not
to make the "obvious" blunders that slow things down. So my immediate
reaction is that such a course could well be very useful for somebody
who implement DSP algorithms.

Having hands-on experience with low-level assembly programming is a
definitive advantage for somebody who works with computer programming,
DSP or not. Somebody with this type of experience would probably do
a better job than somebody without it, given the oportunity to
demonstrate
their capacities. Whether one gets and advantage during the *search*
for 
a job, well, that's a completely different question...

Rune

"kiki" <lunaliu3@yahoo.com> writes:

> --------------HAPPY NEW YEAR TO ALL DSPERS ---------
> 
> Hi all,
> 
> I am considering taking a class in DSP with emphasis on TI processors.
> 
> The course description states: "This course introduces the fundamentals of 
> applied digital signal processing by implementing a wide range of digital 
> signal processing (DSP) applications on the Texas Instrument (TI) TMS320C54x 
> DSP processor."
> 
> It will surely improve my understanding of DSP in general. But it also 
> involves the workload of TMS DSP programming, although my interest is more 
> on algorithm design...
> 
> What do you think? Can DSPer in this forum give me some experiences? Will 
> this help me find an summer intern/job more easily?

You sound ambivalent. Do you or do you not want to work at the
assembly language / implementation level? If not, then don't
bother. I've known several PhDs who are "algorithm guys" that wouldn't
spend 5 minutes thinking about implementation issues. If you choose this
route, then I'd also say your road is more difficult since there are a 
number of very sharp "algorithm guys" in industry and academia. 

But yes, I think the class will help you find a job doing implementation.
-- 
Randy Yates
Sony Ericsson Mobile Communications
Research Triangle Park, NC, USA
randy.yates@sonyericsson.com, 919-472-1124

"Randy Yates" <randy.yates@sonyericsson.com> wrote in message 
news:xxpekh1tgdo.fsf@usrts005.corpusers.net...
> "kiki" <lunaliu3@yahoo.com> writes:
>
>> --------------HAPPY NEW YEAR TO ALL DSPERS ---------
>>
>> Hi all,
>>
>> I am considering taking a class in DSP with emphasis on TI processors.
>>
>> The course description states: "This course introduces the fundamentals 
>> of
>> applied digital signal processing by implementing a wide range of digital
>> signal processing (DSP) applications on the Texas Instrument (TI) 
>> TMS320C54x
>> DSP processor."
>>
>> It will surely improve my understanding of DSP in general. But it also
>> involves the workload of TMS DSP programming, although my interest is 
>> more
>> on algorithm design...
>>
>> What do you think? Can DSPer in this forum give me some experiences? Will
>> this help me find an summer intern/job more easily?
>
> You sound ambivalent. Do you or do you not want to work at the
> assembly language / implementation level? If not, then don't
> bother. I've known several PhDs who are "algorithm guys" that wouldn't
> spend 5 minutes thinking about implementation issues. If you choose this
> route, then I'd also say your road is more difficult since there are a
> number of very sharp "algorithm guys" in industry and academia.
>
> But yes, I think the class will help you find a job doing implementation.
> -- 
> Randy Yates
> Sony Ericsson Mobile Communications
> Research Triangle Park, NC, USA
> randy.yates@sonyericsson.com, 919-472-1124

Thanks a lot Randy!

Really this class is no use? Could you please specify what are those 
"algorithm guys" doing in industry typically? What are those "implementation 
guys" doing? Which type of guys get job more easily? Which has more fun in 
working?

It is clear that I like mathematics and algorithms. But do "algorithm guys" 
never care about implementation details? I guess the people you are talking 
about are those with very specific job classification in large companies... 
In those companies, different types of people do different types of 
things... 

"Rune Allnor" <allnor@tele.ntnu.no> wrote in message 
news:1104845921.286384.167610@c13g2000cwb.googlegroups.com...
>
> kiki wrote:
>> --------------HAPPY NEW YEAR TO ALL DSPERS ---------
>
> Thanks and likewise.
>
>> Hi all,
>>
>> I am considering taking a class in DSP with emphasis on TI
> processors.
>>
>> The course description states: "This course introduces the
> fundamentals of
>> applied digital signal processing by implementing a wide range of
> digital
>> signal processing (DSP) applications on the Texas Instrument (TI)
> TMS320C54x
>> DSP processor."
>>
>> It will surely improve my understanding of DSP in general. But it
> also
>> involves the workload of TMS DSP programming, although my interest is
> more
>> on algorithm design...
>>
>> What do you think? Can DSPer in this forum give me some experiences?
> Will
>> this help me find an summer intern/job more easily?
>
> Well, I don't know what other experience you have with computer
> programming, so I'll try to be careful not to be too definitive in
> my advice. I don't have much experience with DSP programming as such,
> but I did learn a lot of assembly programming on the then state-of-
> the-art i8086 - 80286 + x87 maths processors in the late 80ies/early
> 90ies.
>
> I've hardly done assembly programming after taking those classes,
> but the experience and insights are invaluable. I know a lot more
> about what happens inside the computer, what to do and what not to do,
> what makes a program fast and what slows it down. These days I don't
> do much manual optimization to speed things up, but I do my best not
> to make the "obvious" blunders that slow things down. So my immediate
> reaction is that such a course could well be very useful for somebody
> who implement DSP algorithms.
>
> Having hands-on experience with low-level assembly programming is a
> definitive advantage for somebody who works with computer programming,
> DSP or not. Somebody with this type of experience would probably do
> a better job than somebody without it, given the oportunity to
> demonstrate
> their capacities. Whether one gets and advantage during the *search*
> for
> a job, well, that's a completely different question...
>
> Rune
>

Thanks a lot Rune!

Yeah I have done some INTEL 386+ assembly, but nowadays nobody cares about 
those small details/tricks anymore... People talk more about Java, etc. 
Maybe that's the trend. Maybe we should do Java for DSP chips, instead of 
those low level TI languages...

Yet you have not answered my key question yet...  :=)

>> Whether one gets and advantage during the *search*
>> for a job, well, that's a completely different question... 

kiki wrote:
> "Rune Allnor" <allnor@tele.ntnu.no> wrote in message
> news:1104845921.286384.167610@c13g2000cwb.googlegroups.com...
> >
> > kiki wrote:
> >> --------------HAPPY NEW YEAR TO ALL DSPERS ---------
> >
> > Thanks and likewise.
> >
> >> Hi all,
> >>
> >> I am considering taking a class in DSP with emphasis on TI
> > processors.
> >>
> >> The course description states: "This course introduces the
> > fundamentals of
> >> applied digital signal processing by implementing a wide range of
> > digital
> >> signal processing (DSP) applications on the Texas Instrument (TI)
> > TMS320C54x
> >> DSP processor."
> >>
> >> It will surely improve my understanding of DSP in general. But it
> > also
> >> involves the workload of TMS DSP programming, although my interest
is
> > more
> >> on algorithm design...
> >>
> >> What do you think? Can DSPer in this forum give me some
experiences?
> > Will
> >> this help me find an summer intern/job more easily?
> >
> > Well, I don't know what other experience you have with computer
> > programming, so I'll try to be careful not to be too definitive in
> > my advice. I don't have much experience with DSP programming as
such,
> > but I did learn a lot of assembly programming on the then state-of-
> > the-art i8086 - 80286 + x87 maths processors in the late
80ies/early
> > 90ies.
> >
> > I've hardly done assembly programming after taking those classes,
> > but the experience and insights are invaluable. I know a lot more
> > about what happens inside the computer, what to do and what not to
do,
> > what makes a program fast and what slows it down. These days I
don't
> > do much manual optimization to speed things up, but I do my best
not
> > to make the "obvious" blunders that slow things down. So my
immediate
> > reaction is that such a course could well be very useful for
somebody
> > who implement DSP algorithms.
> >
> > Having hands-on experience with low-level assembly programming is a
> > definitive advantage for somebody who works with computer
programming,
> > DSP or not. Somebody with this type of experience would probably do
> > a better job than somebody without it, given the oportunity to
> > demonstrate
> > their capacities. Whether one gets and advantage during the
*search*
> > for
> > a job, well, that's a completely different question...
> >
> > Rune
> >
>
> Thanks a lot Rune!
>
> Yeah I have done some INTEL 386+ assembly, but nowadays nobody cares
about
> those small details/tricks anymore... People talk more about Java,
etc.
> Maybe that's the trend. Maybe we should do Java for DSP chips,
instead of
> those low level TI languages...

You forgot matlab... knowing assembly languages and having experience
with that sort of programming provides a basis for evaluation when
JAVA or MATLAB will do, and when heavy-duty low-level programming
is required. Way too many people have only written a couple of matlab
scripts, and live on in the delusion that they know how to program.
As I said, experience with assembly languages prepares for system
performance evaluation in ways matlab or other languages never can
challenge.

> Yet you have not answered my key question yet...  :=)

I did not, and that was the intention on my part. I wouldn't dream
of trying to come up with a specific answer for your question.
This sort of experience could well be the deciding factor if I
ever were to choose between two candidates, to a large extent
because I don't have any DSP-processor programming experience myself.
If I ever were to hire somebody, I would be reluctant to hire someone
with no assembly programming experience.

But those are my personal views, based on my own experience. Somebody
else might think in completely different ways. In general, the more
DSP you know (practice/theory, system implementation/use, algorithm
design/programming) the better chance you would have in the
job market. But as you suggested, and others agreed on, these are
judgements you have to make yourself, based on your interests and
the amount of work or effort you are willing to spend on learning
a new skill.

Rune

> >> Whether one gets and advantage during the *search*
> >> for a job, well, that's a completely different question...

"kiki" <lunaliu3@yahoo.com> wrote in message
news:crev8t$7ks$1@news.Stanford.EDU...
>
> Really this class is no use? Could you please specify what are those
> "algorithm guys" doing in industry typically? What are those "implementation
> guys" doing? Which type of guys get job more easily? Which has more fun in
> working?

For me, the ideal job is a blend of both algorithms and implementation details,
probably leaning more toward implementation/"making it work".  Of course, this
is just a personal preference and others may have a completely different
opinion.  Personally, I get a real kick out of optimizing assembly code
line-by-line, but some might consider that a disease!  :-)

"kiki" <lunaliu3@yahoo.com> writes:

> "Randy Yates" <randy.yates@sonyericsson.com> wrote in message 
> news:xxpekh1tgdo.fsf@usrts005.corpusers.net...
>> "kiki" <lunaliu3@yahoo.com> writes:
>>
>>> --------------HAPPY NEW YEAR TO ALL DSPERS ---------
>>>
>>> Hi all,
>>>
>>> I am considering taking a class in DSP with emphasis on TI processors.
>>>
>>> The course description states: "This course introduces the fundamentals 
>>> of
>>> applied digital signal processing by implementing a wide range of digital
>>> signal processing (DSP) applications on the Texas Instrument (TI) 
>>> TMS320C54x
>>> DSP processor."
>>>
>>> It will surely improve my understanding of DSP in general. But it also
>>> involves the workload of TMS DSP programming, although my interest is 
>>> more
>>> on algorithm design...
>>>
>>> What do you think? Can DSPer in this forum give me some experiences? Will
>>> this help me find an summer intern/job more easily?
>>
>> You sound ambivalent. Do you or do you not want to work at the
>> assembly language / implementation level? If not, then don't
>> bother. I've known several PhDs who are "algorithm guys" that wouldn't
>> spend 5 minutes thinking about implementation issues. If you choose this
>> route, then I'd also say your road is more difficult since there are a
>> number of very sharp "algorithm guys" in industry and academia.
>>
>> But yes, I think the class will help you find a job doing implementation.
>> -- 
>> Randy Yates
>> Sony Ericsson Mobile Communications
>> Research Triangle Park, NC, USA
>> randy.yates@sonyericsson.com, 919-472-1124
>
> Thanks a lot Randy!

OK, well maybe I spoke too soon with that "smart-ass" remark. Sorry -
I hadn't read this post yet. I apologize. I tend to take usenet posts
the wrong way sometimes.

> Really this class is no use? 

Well, no - not if you're goal is to obtain a PhD and work in
algorithms all day. Most of the PhDs I know that do this seldom get to
the assembly language level.

> Could you please specify what are those 
> "algorithm guys" doing in industry typically? 

Writing papers, going to conferences, and developing Matlab
code. Occasionally they contemplate their navel as well.

> What are those "implementation guys" doing?

Optimizing C or assembly code, moving things around in the memory
map to make everything fit, writing trouble reports, documenting
the implementation, reading through someone else's code through
which you must interface, etc.

> Which type of guys get job more easily? 

Couldn't really tell you. My gut is that the PhDs have a harder
time landing the ideal "algorithmic ivory tower" jobs.

> Which has more fun in 
> working?

Yeah, well I'm pretty sure you know I don't have the anser to
that since it is subjective.

> It is clear that I like mathematics and algorithms. But do
> "algorithm guys" never care about implementation details? 

Not never, but seldom to the level of assembly programming. 

> I guess the people you are talking about are those with very
> specific job classification in large companies...  In those
> companies, different types of people do different types of things...

Maybe.
-- 
%  Randy Yates                  % "She's sweet on Wagner-I think she'd die for Beethoven.
%% Fuquay-Varina, NC            %  She love the way Puccini lays down a tune, and
%%% 919-577-9882                %  Verdi's always creepin' from her room." 
%%%% <yates@ieee.org>           % "Rockaria", *A New World Record*, ELO   
http://home.earthlink.net/~yatescr

"kiki" <lunaliu3@yahoo.com> wrote in
news:crevif$81r$1@news.Stanford.EDU: 

> 
> "Rune Allnor" <allnor@tele.ntnu.no> wrote in message 
> news:1104845921.286384.167610@c13g2000cwb.googlegroups.com...
>>
>> kiki wrote:
>>> --------------HAPPY NEW YEAR TO ALL DSPERS ---------
>>
>> Thanks and likewise.
>>
>>> Hi all,
>>>
>>> I am considering taking a class in DSP with emphasis on TI
>> processors.
>>>
>>> The course description states: "This course introduces the
>> fundamentals of
>>> applied digital signal processing by implementing a wide range of
>> digital
>>> signal processing (DSP) applications on the Texas Instrument (TI)
>> TMS320C54x
>>> DSP processor."
>>>
>>> It will surely improve my understanding of DSP in general. But it
>> also
>>> involves the workload of TMS DSP programming, although my interest
>>> is 
>> more
>>> on algorithm design...
>>>
>>> What do you think? Can DSPer in this forum give me some experiences?
>> Will
>>> this help me find an summer intern/job more easily?
>>
>> Well, I don't know what other experience you have with computer
>> programming, so I'll try to be careful not to be too definitive in
>> my advice. I don't have much experience with DSP programming as such,
>> but I did learn a lot of assembly programming on the then state-of-
>> the-art i8086 - 80286 + x87 maths processors in the late 80ies/early
>> 90ies.
>>
>> I've hardly done assembly programming after taking those classes,
>> but the experience and insights are invaluable. I know a lot more
>> about what happens inside the computer, what to do and what not to
>> do, what makes a program fast and what slows it down. These days I
>> don't do much manual optimization to speed things up, but I do my
>> best not to make the "obvious" blunders that slow things down. So my
>> immediate reaction is that such a course could well be very useful
>> for somebody who implement DSP algorithms.
>>
>> Having hands-on experience with low-level assembly programming is a
>> definitive advantage for somebody who works with computer
>> programming, DSP or not. Somebody with this type of experience would
>> probably do a better job than somebody without it, given the
>> oportunity to demonstrate
>> their capacities. Whether one gets and advantage during the *search*
>> for
>> a job, well, that's a completely different question...
>>
>> Rune
>>
> 
> Thanks a lot Rune!
> 
> Yeah I have done some INTEL 386+ assembly, but nowadays nobody cares
> about those small details/tricks anymore... People talk more about
> Java, etc. Maybe that's the trend. Maybe we should do Java for DSP
> chips, instead of those low level TI languages...
> 
> Yet you have not answered my key question yet...  :=)
> 
>>> Whether one gets and advantage during the *search*
>>> for a job, well, that's a completely different question... 
> 
> 

The regulars on comp.dsp tend to fall into two broad categories: 
generalists & specialists. 

I think you can get very different answers depending on which category  
best fits the respondent.

Learning 386 assembly was important even though nobody really spends much 
time with 386 assembly anymore. I probably have forgotten more languages 
than I still know (maybe 10) and am not really a programmer at all. 

Algorithm guys benefit from learning assembly coding. I also think that way 
too many programmers need to learn a lot more about the theoretical side.

Your question also reminded me of a situation I had many years ago. I hired 
a purely theoretical guy who I had working a noise reduction algorithm for 
speech. I asked him how it was working and he showed me his matlab 
simulation. I said that's fine, how does it sound? He had no idea, he had 
never listened to it! I told him that his matlab program didn't buy our 
products, customer's did and they all had ears.

The point I'm making is that find ways to become a more rounded engineer. 
If you want to be an algorithm guy, thats fine, but make sure you have an 
exposure to other areas. 

So my advise to you is to take the class. It doesn't really matter whether 
its a TI , ADI or Freescale DSP. If you can learn one, you can learn 
another. It will also help make what you've learned so far, a little more 
real.

Will it get you a summer intern job? I think its far more likely someone 
will hire you for a short term programming task, than for more theoretical 
work that may utilize some of the other skills that you are developing. For 
longer term employment, there are opportunities for all kinds of engineers.  

You might guess that I'm one of the generalists, I like to think that I 
know a little bit about just about everything ;-)

  

-- 
Al Clark
Danville Signal Processing, Inc.
--------------------------------------------------------------------
Purveyors of Fine DSP Hardware and other Cool Stuff
Available at http://www.danvillesignal.com

Randy Yates <yates@ieee.org> writes:
> [...]
>> What are those "implementation guys" doing?
>
> Optimizing C or assembly code, moving things around in the memory
> map to make everything fit, writing trouble reports, documenting
> the implementation, reading through someone else's code through
> which you must interface, etc.

Oh, and I forgot to add: Modifying the algorithm to work better
and faster in 1/5th the (labor) time the PhD spent designing it.

Of course this characteristic is not always appropriate. There are
many PhDs who do real work. The question is: what is your goal? 
Personally, I like to have a foot in both worlds.

--RY

-- 
%  Randy Yates                  % "Maybe one day I'll feel her cold embrace,
%% Fuquay-Varina, NC            %                    and kiss her interface, 
%%% 919-577-9882                %            til then, I'll leave her alone."
%%%% <yates@ieee.org>           %        'Yours Truly, 2095', *Time*, ELO   
http://home.earthlink.net/~yatescr