Logo
Forums
Forums
UniSearch
UniSearch
UniSuggest
UniSuggest
AdmitTrend
AdmitTrend
CourseLiX
CourseLiX
RoommateFinder
RoommateFinder
EduTravel
EduTravel
EdulixExpress
EdulixExpress
Chat
Chat
Blog
Grad Studies
About Us
About Edulix

Post Reply  Post Thread 
 
Thread Rating:
  • 34 Votes - 4.5 Average
  • 1
  • 2
  • 3
  • 4
  • 5
The Embedded Systems Thread (Please read first post)
Author Message
Zodiac Away
Edulix MVP
*********
Edulix MVP

Zodiac Away
Edulix MVP
*********
Edulix MVP


Posts: 605
Likes Given: 1
Likes Received: 10 in 8 posts
Joined: Jun 2008
Unisearch: Link
Reputation: 39
Post: #1
The Embedded Systems Thread (Please read first post)
0
0
Hello guys,

PS: I have stopped managing this thread. Please PM me if you have any questions. And before you PM me, read this. If your PM does not conform to the writing etiquettes mentioned here, I will not reply to your questions.


I see a lot of students applying to grad school with interests in embedded systems / VLSI / computer architecture / other related disciplines. The focus of this thread is to try and cover research in embedded systems in various universities.

Before you read on, some very important things to note:

1) This thread might contain information that is stale (or has become stale over a period of time) or I might have missed out on some details somewhere. If you know something more about research work in some XYZ university than what is already been covered in the thread, PLEASE POST YOUR FINDINGS SO THAT I CAN UPDATE THE FIRST POST.

2) Strictly no profile evaluations!!! The objective of the thread is not to see if some university is an ideal fit for you. That is for you to decide based on information available here and/or other sources. There is a profile evaluation forum for this purpose.

3) I will really appreciate it if I could get help in managing this thread. By help, I mean HELP. Not just replying 'I am in' here and then forgetting about it. That makes no sense whatsoever to me. I am open to trying out new ideas to keep things more organised and up-to-date. If you have any new suggestions...Please PM me and don't reply here.


USEFUL VIDEOS ON EMBEDDED:
by Dr. Shantanu, Dept of Electrical Engineeing, IIT Delhi

1. Introduction to Embedded Systems (thanks to esd again for this link)

2. Software for Embedded Systems

3. Embedded Hardware

LECTURES NOTES ON EMBEDDED SYSTEMS FROM ELECTRICAL ENGG DEPT, IIT Kharagpur (thanks to ksr5a0 for providing this link)


OTHER USEFUL LINKS ON EDULIX:
k-electron's thread on universities based on various specializations in ECE:

VLSI focused thread managed by venkatesh and Madmax

UNIVERSITIES IN DETAIL (still not complete):

Arizona State University :
Department : Computer Science and Engineering
Specialization : Embedded software, SoC design (Dr. Chatha), Compiler design & optimization for embedded systems (Prof. Aviral Shrivastava)

Iowa State University :
Department: Electrical and Computer Engineering.
Specialization: reconfigurable embedded systems, real-time embedded systems design, parallel and distributed systems, VLSI

Main Areas:
Software Systems

VLSI

Virginia Tech:
Department: Electrical and Computer Engineering
Specialization: Real-time systems, reconfigurable system design, OS
Main Areas:
Real Time Systems Lab
Configurable computing lab

University of Maryland, College Park:
Department: Electrical and Computer Engineering. Certain profs are also from Department of Computer Science.
Specialization: Embedded software, embedded signal processing, memory management for embedded systems
The link to the Embedded System Design Lab is not active. Instead I am posting links to the profiles of two profs who are working in this field.

Rajeev Barua
Mainly works on compilers, other aspects of embedded software, computer security and also memory management in embedded systems. His current research work is based on reliable software design for embedded systems.

Prof Shurva
Current research work is on embedded signal processing. He has earned Fulbright grant. He has also authored many text books in the same field.

North Carolina State University
Department: Electrical and Computer Engineering
Specialization: ALL. Seems to be tending towards embedded h/w in my opinion.

Carnegie Mellon University
Department: Electrical and Comp Engg
Specialization: ALL. Without doubt the best school for embedded and comp architecture. If you can manage the fees then this should be there on your list as ambi.

Georgia Tech
Department: Electrical and Comp Engg
Specialization: ALL. Among the top schools for embedded and comp architecture.

University of California, San Diego
Department: Electrical and Comp Engg
Specialization: SoC design and other areas of embedded h/w.

University of California, Irvine :
Department: Electrical and Comp Engg
Specialization: ALL. Good school for embedded and computer architecture studies. Especially strong in reconfigurable computing on the comp architecture side.
Research Groups:
Centre for Embedded Systems
Profs like Frank Vahid , who are authors of popular text books on embedded systems, are a part of this group.
Advanced Computer Arch Group

University of California, Santa Barbara
Department : Electrical and Comp Engg
Specialization : parallel and distributed systems, SoC design and other areas on embedded h/w.

Penn State University
Department: Comp Science and Engg
Specialization: ALL. There are three different labs and between them they cover various research themes. IMO, stronger in embedded h/w.

University of Penn
Deparment: Dedicated program for embedded systems that falls under the purview of the Computer and Info Science and Electrical and Systems Engineering departments.
Specialization: ALL. Active research work going on in wireless sensor networks area.

TAMU (thanks to jagadishc for providing me these links)

profs and labs
==========
Rabi N. Mahapatra
Embedded Systems and Codesign Laboratory
>>>seems to be concentrated on reconfigurable architectures and networks, some real time systems

Eun Jung Kim
High Performance Computing Lab
publications
>>> seems to be parallel and distributed computing(i think)

Valerie Taylor
publications
>>> seems to be parallel and distributed computing

Radu Stoleru
Embedded Networked Sensor Systems
publications
>>> wireless networks

Lawrence Rauchwerger
publications
>>> parallel computing, compiler design

courses
======
though i don't know which are all actually offered.

BOOKS:
Embedded Software:
For beginners:
1)An Embedded Software Primer: David E Simon
The starting half of the book is a bit of a waste in my opinion. It covers hardware for computer science folks. But I think it is too basic and can be read from any other source.The second half is good for those who have no prior background to RTOS and embedded programming. This book explains the common problems encountered in programming for real-time systems. The lead up to why exactly an RTOS is needed is very nicely explained. Probably not for those who have prior experience in working with some RTOS. This book is very easy paced and language used is very easy to follow.

2)Operating Systems Principles : Galvin, Gagne, Silberschatz
For those who have no background in operating systems, this is a good book to start with.

3)Embedded C: Micheal Pont
C Programming using Keil and 8051 as example. He covers important aspects on programming in embedded C. Very good book for people from computer science background who want to get their hands dirty with working a microcontroller and a cross compiler.

4)Embedded System Design - Frank Vahid & Tony Girvargis (thanks to esd for bringing this to my notice and thanks to ksr4a0 for providing the link to the text)
Quite useful for introduction to memory management in embedded systems (like cache mapping techniques),types of memory (sram, dram),and also other protocols (i2c,spi,can,usb,firewire). Good for one-time read but not worth keeping as reference.

For experienced:
1)Embedded Linux Design and Development: P Raghavan
Certain concepts may be a little difficult for the beginner to grasp. Pretty decent book. There are times when he digresses from the main topic

For all:
1)Linux Kernel Development: Robert Love
Good book on kernel internals. Must read for someone who is
interested in working in embedded linux, kernel programming, RTOS. Easy
to follow for a complete newbie.

2)Linux Device Drivers: O'Reilly
Perhaps the only book there is on the subject. This is a bible for those
interested in working on device drivers. You need to have some background
in Make and makefiles. You also need to work out the examples and try
it yourself to get the best out of this book (I mean this! I know this can be
said about every other book but this is a MUST for this book).
Very vast and comprehensive.

3)C Programming FAQs : Steve Summit
This is a very very good book to clear all your doubts in C. I have sometimes also found useful programming tips from this book, related to programming style, debugging techniques etc. Worth buying and keeping as a reference.

Embedded Hardware:

1)Designing Embedded Hardware : O'Reilly
Covers the basics well. Not that useful for someone who has prior
experience with working on certain design problems in embedded h/w.

For other books, please refer to the VLSI focus thread for now. They have discussed popular books on Soc design, ASIC, digital VLSI etc.

Books NOT worth reading:
1)Fundamentals of embedded software- where C and assembly meet: Daniel Lewis
Presentation is quite poor and confuses you more.

2)Embedded Systems Architecture: Tammy Norgaard
Although the organisation is ok, the actual content is junk. For embedded h/w, he explains stuff like p-n diode and its functionality, resistor colour code and what not. That is way too basic and not worth buying the book for.

3) Embedded Hardware- Jack Ganassle
Don't buy this book!

4) Embedded System Design- Steve Health
Lot of elementary mistakes. Content is also not that great.

PENDING ACTION ITEMS:
1. Add more universities to the above list.

2. Prepare a comprehensive document introducing the various fields in embedded systems to people who are not familiar with this domain.

Any help in managing this thread would be appreciated! PM me if you are interested Smile

________________________________________________________________________________​_____________________

VLSI Specific Documentation by Pramod Natraj

Specialisations In VLSI

1) Analog/Mixed-signal - this field is one of the most popular areas falling under VLSI. And it will stay popular, as long as human beings perceive information in the analog way. Analog circuits mostly deal with, as the name suggests, analog signal processing. Egs, Amplifiers, PLLs, Oscillators, etc.

Mixed-signal is kind of a sub-branch of analog. As now most standard chips have a DSP core, mixed-signal blocks, being at the i/p nd o/p, form one of the most important blocks on the chip. They define the final specifications of the chip. Hence, these will never go out of fashion.

2) RF - This is what the microwave engineers call their work nowdays. At RF freqs, though the some analog pinciples apply, parasitics, in both active and passive devices, start playing a helluva important role. RF again involves two aspects - device research and receiver/other circuits design, which again can be further classified depending on what kind of substrate you want to work on. Has gradually become one of the most popular fields all over, as engineers try to push the freq limits, while accommodating multiple standards on the same chip. Last I heard was a transceiver designed at 60 GHz!!

3) Digital - The DSP core that every chip consists of needs to be designed as well. Again a good field to work in. Imp research includes, power efficient DSP algorithms to implement features like FFT, etc, low power architectures, ASIC and FPGA design, leakage reduction, etc.

4) VLSI - This deals with issues involved in chip design and manufacture. Imp research going on - interconnect/link design, chip architecture optimization, 3D chips, etc.

5) VLSI CAD - This involves the design of tools which engineers use to design chips. Almost every major univ has a research group in this field - serves to underline the importance of this field. Again research includes algorithms, AI, etc.

6) Fault Detection and Testing - Now that the chip is ready, one needs to make sure that it works perfectly, before shipping it. Here fault detection tools come in handy. Digital chips can be easily tested coz of binary nature of their output, but analog chips are more tricky. Imp research includes - Built-In Self Testing (BIST), etc. I know of 2 grps at Auburn and Iowa state, working on Fault detection.

Some other specialized avenues are opening up - such as biomedical/bio-inspired/neuromorphic VLSI. These are still new fields, and if one wishes to work in them, he or she can still make an impa

Universities and VLSI Specialisation
OREGON state university-- This is an excellent place for Analog VLSI, and what with the close proximity of semiconductor industries, post MS/PhD jobs would be easy. I'd rate this as very very ambitious.

ASU--Was impossible to get thru for fall 08 batch. This, I believe was mostly due to the fact that they took too many students during the last fall. So, this mite mean they won't be as strict for the coming fall. Just my assumption

UMass, Amherst--Very very good place for Digital VLSI. Having relevant work experience helps here.

OSU-- NO WORK GOING ON IN DIGITAL VLSI. Its a good place for analog. students from VTU, don't bother applying.

UCSB-- Good work in Digital as well as analog vlsi. I'd rate this as ambitious looking at the current trend.

TAMU-- very good for mixed signal. Very erratic with their admit.

UT Austin- Excellent place for VLSI, any discipline. Though they did give quite a lot of admits for Computer Engg stream, some really good profiles got rejects. So this still stays ambitious.

Gatech- The faculty strength in computer engineering in specific, ECE in general, is breathtaking-ly Huge... Plenty of opportunities in any stream of VLSI.

UMinn twin cities . This place is pretty selective. Good for digital vlsi.

UFL-- good work in digital VLSI. Lots of desi's though!

Iowa state university-- few prof's there are doing good work in dig. vlsi. Very few intl intake

UMCP-- very good for analog vlsi. very very tough to get in.

UCLA-- I'd rate this as the best place to be if you're into analog vlsi. Amazing profs. Amazing facilities. Very tough to get in.

Stanford Uni- 'nuff said

NCSU-- digital VLSI research is supposed to be good here.

Penn State-- This place is supposed to be on par with UMn and OSU as far as the VLSI research is concerned.

Virginia Tech-- Good place for digital vlsi. They offer Meng to most MS applicants.

Duke University-- Small Dept. but good research in vlsi.

UoA-- again, small dept. Not so great for comp. engg.

Rutgers-- Not so great for VLSI.

SUNY- SB-- I've heard vlsi research is almost dead here.

paragk:
1. Berkeley - obviously, this is where a lot of VLSI-CAD originated. Albertro Sangiovani Vincitelli is here, he is credited by many as the father of modern VLSI-CAD. Incredibly difficult to get into. They look for a publication if you are not a IIT topper. Design Automation Conference (DAC) or Asia-Pacific DAC, or else no chance.

2. Minnesota Twin Cities - not that famous, but there is VERY strong VLSI-CAD (and VLSI design) research group there. Prof. Sachin Sapatnekar, look him up. Very famous for recent contributions to logic and physical synthesis. The alumini from this group is littered across Intel, Cadence, Synopsys and they have EXCELLENT tie-ups for internships. If you are from a good college its possible to get into the MS program. The PhD program directly is almost impossible for non-IIT/BITS. (exceptions obviously, if you have awesome publications)

3. UCLA - EDA/VLSI-CAD labs are there in both CS and EE departments. Prof Chong in CS and Prof He and Prof Gupta in EE are all amazing. And have very very high industry standing. This is a group that is willing to substitute relevant work-ex for publications for the MS program. Not so for the PhD - you have to be IIT + publications/internship.


4. UTAustin - Prof. David Pan. Quite famous for "timing driven placement" algos. Very very good course structure and again good industry recognition. Everyone knows what it takes to get into UTexas Austin (and I didnt have it ) so I wont say anything more.

(from this point onwards the rankings/ratings are a little hazy, that is they are all more or less equally good)

5. Michigan Ann Arbor - prof Ivan Markov, good work in physical synthesis.

6. UC San Deigo - Prof Khang's group. No idea what it takes to get in. But it is awesome, a lot of cutting edge work.

7. USC - people know enough about this college I guess. Prof. Pedram is quite good, low-power design and CAD.

8. Univ of Wisconsin - a relatively small but excellent CAD group. This one makes it to "top" (in my mind atleast) because they are in Wisconsin ,which is where Mentor Graphics is from and they have excellent tieups with the group here.

9.TAMU - is actually mind blowing for VLSI-CAD. But I heard they are selective and look for IIT/big-univ tag (the CAD group I am talking about). I personally know a LOT of people at Cadence who are from TAMU and are considered to be some of the best brains at Cadence. For the MS program I have seen that they are favorable/partial to Pune and Bombay University.

10. Irvine - again not so well known but the ACES group/lab is making a good niche for itself. Mostly CAD and architectures for embedded, Prof. Sunil Dutt (ex-BITS) is the professor here. He is quite keen on Indian students (he came to my college and gave a lecture once, and he said so ), though the group seems to be partial towards Delhi Engineering colleges.

Finally a list of colleges that are not the "big names" but still have a decent VLSI-CAD group -

1. SUNY Binghamton - Prof. Patrick Madden
2. Univ of Cincinati - dont remember the Profs name. But the group is decent, a lot of the alumini are in excellent places like Synopsys and Intel.
3.Boulder Colarado - strong research in logic synthesis, which originated at this place + berkeley.
4. Colarado State Univ - again decentish research. Easy to get in. Prof. Sudeep Parsicha is excellent.

Outside US -
1. Toronto - Prof Brown, excellent group. Hard to get into though is what I gather.
2. Univ of Aachen Germany
3. Enrico Macii at Univ of Turin Italy.

I hope this helps in the document that you are compiling. Not many people are really interested in VLSI-CAD, at least not before they go to the US, but hope it helps a few guys at least.

Analog VLSI
Analog VLSI mostly deals Transistor level circuit design, fault tolerance testing etc.

Digital VLSI
the algorithms and software tools for CAD/Electronic Design Automation, design based on FPGA (including reconfigurable computing)...etc

digital VLSI, as i mentioned before, deals with digital integrated circuits. By architecture, i meant digital chip architecture. Not much coding is involved here. E.g. - Prof. Chandrakasan's group at MIT - http://www.rle.mit.edu/rleonline/researc...Group.html

Putting it bluntly, Digital VLSI aims to design a single digital chip. Whereas, comp arch utilizes a number of chips to optimize processing. This is a very very crude definition

Computer Architecture(2 levels of higher abstraction than Digital VLSI)
http://www.ece.ncsu.edu/research/cas/
http://lca.ece.utexas.edu/
Computer architecture is the engineering of a computer system through the careful design of its organization, using innovative mechanisms and integrating software techniques, to achieve a set of performance goals.
The most common goals in computer architecture revolve around the tradeoffs between cost and performance (i.e. speed), although other considerations, such as size, weight, reliability, feature set, expandability and power consumption, are important factors as well.

Architecture designs also includes the system on chip (SoC), system on programmable chip(SoPC) etc. You make use of the EDA tools to design the system. You'll be coding using the HDL's. Reconfigurable computing, dynamically reconfigurable architectures are prolly the hottest research topics in these fields.




Networks and Communications
I donno how much ppl are interested in communications, But let me pitch in as well.
There is a fundamental difference between networks and communications. Right now let me take wireless, wired, Point to point, Point to multi point, Adhoc, Multiple access networks etc....
Communications in general deals with the lowest level in a network protocol. You guys remember different layers in network models right?
Communications deals with the physical, MAC, sometimes network layer as well.
Whereas Networks generally deals with higher layers : presentation, application, Transport, etc etc....
The subjects that you generally study for communications are :
Information theory, Wireless Communications, Estimation and Detection(though its also signal processing), Stochastic process, Mobile communications, Digital communication, Channel coding, Error control coding, Satellite communication. Etc
Subject that you encounter in networks are:
Wireless networks, Adhoc Networks, Network performance, Computer networks, Mobile data networks, Lan, Wan, Man, (CDMA,TDMA)Multiple access networks, Data networks etc etc

If you study network courses your job description would be network programmer/engineer.
Study communication courses your job description would be Network designer/engineer/developer.

Generally ppl studying communications have more scope for research (or ppl who wanna do phd are gonna take these courses).
and PPl who wanna go for a job take the Network courses.

Most network courses are offered in univs that specialize in Telecommunication and wireless communication courses.
these univs I need not mention.

but univs which have good research in communication are as follows (most of them may be super ambis but anyways) :

UCSD, stanfy, Columbia, Uwisconsin, Uminn Tc, U wash seattle, Vtech, Purdue, USC (one of the biggest commn depts), UCLA, UCSB, UCI, UFL(surprisingly pretty good), GATECH, NCSU(eventhough network courses are better here).

The typical coursework that will cover the breath of embedded systems. I am listing the areas you need to know, (these are not specific to any university) :

1. Computer Architecture
2. Systems Level Design (Embedded)
3. FPGAs
4. Parallel Processing
5. Distributed and High Performace Computing
6. Operating Systems
7. Algorithms (people dont often realize how important this course is for an embedded enginner, not until they get a full time position)
8. Real Time Operating Systems
9. Very good working knowledge of C and C++ (or JAVA)
10. Digital VLSI (just to make sure you know the system (at the lowest level of abstraction) you are dealing with)

Now if you have knowledge in most of the above areas you can call yourself an embedded engineer. And you will surely get a full time position.
Any university you are interested in check to see if there are courses in the above listed areas.

List of Books:
Books
This is just an excerpt from what I have written so far in the document( have just mentioned the name of the Authors):

Digital

1) Weste and Harris (Digital circuit design- excellent book though if you can get the 2nd edition that will be better than the third one)
2) Rabaey (Digital circuit design- Nicely complements the Weste's book on certain issues)
3) Samir Palnitkar (Verilog)
4) Nawabi (VHDL)
5)Digital design using vhdl/verilog by Prof. Stephen Brown (ECE univ of toronto)

Analog

1) Razavi (CMOS based)
2) Gray and Meyer (CMOS, BJT and BiCMOS)
3) Ken and Martin (Has better chapters on Mixed Signal design compared to the others)
4) Philip Allen et al (One with better fundaes but is not a preferred textbook in many univs. Slightly higher level and hence a good hand book)
5) Jim Williams (The art and science of Analog circuit design- Excellent reference book once you work)


RF

1) Behzad Razavi (RF microelectronics - very succinct which is also its biggest flaw. One of the most popular text book)
2) Thomas Lee (RFICs- I believe that this is the best book as it goes like a story and is a perfect bedroom/bathroom companion)
3) Steven Cripps (His books are for serious RF practitioners- His books are the best for Engineers in Power Amplifiers for RF)
4) Asif Abidi (He is a Prof. in UCLA and he has been behind all the seminal work on wireless circuit design. Read his papers. I do not believe he has a book)

Devices

1) Streetman & Banerjee (The most popular book on devices. Very important from interview point of view)
2) Tsividis (Best book on MOSFET in particular. Too much in detail but reading it is worth is as it covers all possible information on MOSFET)
3) Peter Aean et al (Device modeling. One of the best book on device modeling for RF devices in particular. Main focus is on LDMOSFETs. The author is working with freescale industries)

Basic Electronics

1) Sedra and Smith (De facto book on basic electronics. Must book for any engineer)
2) Behzad Razavi (This book just came out last year. Nothing new information but those who like Razavi's style of teaching will love this book. I consider this much better than Sedra and Smith)
3) Art of Electronics (This is a timeless classic which is the most practical book on circuit engineering)
4.Op Amps for Everyone by Ron Mancini (From Texas Instruments)
5. Boylstead

Sedra Smith--> Analog CMOS (Razavi)-->Allen and Holberg

The way ahead-TESLA
What do you actually mean by VLSI? The term VLSI is used to represent soo many domains which by themself are very vast fields.. It all depends on the what you mean..

a) If you mean Analog Circuit design.. Then go by Sedra smith then Razavi and then advanced books on topics like Data converters..

b) If you mean Digital design and front end then go by Moris Mano and read Verilog or VHDL and do projects in it.. (Am sorry abt digital books as I don't have much knowledge in them)

c) If you mean Physical design (Backend) in digital design, then in addition to reading basics through Moris mano, go for Rabey and then do some projects in back end.. If you have access to Cadence SoC encounter, it will be a very big advantage.

d) If you mean Computer Architecture - mmm no idea over the books. Some comp arch senior can help in here..

e) If you mean RFIC design, then in addition to books to be read for Analog, add Tom Lee's RFIC design book.. Razavi sir also has written a book on this.. But since I used only Tom Lee I can confidently recommend that.. Awesome book..

f) If you mean Devices, again no idea in that front.. Streetman, Tsividis books are some I have read or heard good in this area.. Refer Venkatesh's list of books..

g) If you mean VLSI CAD.. Then apart from basic digital design knowledge, you would require strong skills in Linear Algebra I think.. Try Gilbert strang for that. Thats what my friends used and I have no other idea over CAD..

Hmmm so what am trying to mean at the end is.. VLSI is being used for soo many vast field and it is very unfair to simply ask tell me a basic book for VLSI..

Both Verilog and VHDL is being used in industry.. But I have seen Verilog being used more widely than VHDL in industry.. Many use VHDL as well.. I think being strong in either of these will do.. Just knowing basics of other language should be fine..


Venky’s Corner

See, Bolystead is not used in US in top univs. Only Sedra and Smith is used. So, if you wish to get tuned to US style of learning circuit analysis, Sedra's book will be better. The thing is some profs while interviewing you for a RA/TA position may ask you which books you have studied. It is better to state the common books as they can measure the strength of your academic pedigree.

Reg. Verilog book, Micheal Ciletti's book is also excellent. I used it during my final year project but then it is again kinda more complex. It is also not used as a standard textbook in many univs. All the books that I had mentioned are those which are the most popular textbooks at different univs.

See, no one teaches verilog specifically here in the classroom. They just distribute some reference notes. Students can choose what ever book they wish to use. Many Americans as well as international students tend to choose the simplest one as it will generally cover all the basics. That's why Palnitkar's book is kinda popular. Moreover, VHDL is the preferred HDL in academia and many courses will use it. Verilog, on the other hand, is used by the industry where each company has its own manuals and handbooks which are used for training. All these books on verilog/vhdl are then used only as a reference. And that's where Palnitkar's book is good. Its like those MBD guides where everything is simplified with basic examples.

on jobs:
Digital based jobs are the easiest to get and also the pay is the least in the VLSI sector
Mixed signal based jobs are also abundant but they require good projects (FPGA based)
Analog jobs are driven by the market demand and the getting jobs is based on contacts.
RF is the most difficult of all and at the same time, the highest paid. 100% based on recommendations and the groups you work with.
Devices are relatively easier to get, but you should be a PhD student. Otherwise, difficult to get.
It is common to have 65-70K for Digital guys and 75-80K for Analog. Rf can touch 85-90K to start with. I know cases where RF PhD guys got 100-120K to start with.


TOOLS
Tools:Venkatesh
hspice is better thing to learn but then I do not know if many univs in India have access to it. Pspice is not a bad alternative. You will mainly use Cadence here with its built in simulator spectre.

Linux, perl


Tools:Tesla
For Digital front end design: VHDL, Verilog. Recently languages like System Verilog are also being used.
Digital back end: Scripting languages like Perl, Tcl

General system level design: MATLAB

CAD: Scripting languages and also Verilog, VerilogA, VHDL, System Verilog and the likes.

FPGAs: Verilog, VHDL

Useful Websites:
1. http://www.edaboard.com
2. http://www.mosis.org

Source of Information: Thanks to all the inputs from seniors tesla,paragk,venkatesh and other seniors. This information is taken from the archived threads on edulix.

P.S: If there is any inconsistency please let us know so that it can be corrected.

--
Pramod Natraj
________________________________________________________________________________​_____________________

NOT very active on edulix these days. For any questions or help, PM me (expect delayed responses).
Please do NOT tag me for profile evaluations.
(This post was last modified: 05-16-2013 01:27 AM by The Black Swan.)
03-22-2009 05:30 PM
Find all posts by this user Like Post Quote this message in a reply
The Madmax Offline
Edulix MVP - 10K Posts
*********
Edulix MVP

The Madmax Offline
Edulix MVP - 10K Posts
*********
Edulix MVP


Posts: 10,106
Likes Given: 0
Likes Received: 23 in 15 posts
Joined: Nov 2004
Unisearch: Link
Reputation: 178
Post: #2
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
ApplauseApplauseApplause

As if I give a damn
03-23-2009 03:11 AM
Find all posts by this user Like Post Quote this message in a reply
sanjay.devnani Offline
Edulix Senior Member
******

sanjay.devnani Offline
Edulix Senior Member
******


Posts: 123
Likes Given: 0
Likes Received: 0 in 0 posts
Joined: Sep 2008
Unisearch: Link
Reputation: 2
Post: #3
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
Dude!! Great initiative! Applause


I had a few queries:

I am from EE background, and a fall 09 applicant! got into USU and Univ of utah. One of which is more of embedded control systems and the other's robotics. more of CS robotics. Which i presume will also have a lot of embedded technology.

What do u think is better considering ONLY the job prospects.

03-23-2009 12:20 PM
Visit this user's website Find all posts by this user Like Post Quote this message in a reply
Zodiac Away
Edulix MVP
*********
Edulix MVP

Zodiac Away
Edulix MVP
*********
Edulix MVP


Posts: 605
Likes Given: 1
Likes Received: 10 in 8 posts
Joined: Jun 2008
Unisearch: Link
Reputation: 39
Post: #4
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
(03-23-2009 12:20 PM)sanjay.devnani Wrote:  Dude!! Great initiative! Applause


I had a few queries:

I am from EE background, and a fall 09 applicant! got into USU and Univ of utah. One of which is more of embedded control systems and the other's robotics. more of CS robotics. Which i presume will also have a lot of embedded technology.

What do u think is better considering ONLY the job prospects.

Both embedded control systems and robotics are closely related fields. These fields that you are talking about will strengthen your understanding of hardware a little bit more than software.

By CS robotics, I think you are saying that your work will be on the software side of robotics i.e more of coding, understanding real-time systems, scheduling and stuff.

Which one to go for? Hmmm ....thats a tough one. It really depends on what kind of work you are interested in. If you want a good exposure to both the hardware as well as the software side, I would say go for the second option. Of course, you would have to see what companies are present in an around Utah so that you can target them accordingly. By choosing the second option, you are open to opportunities for application development on the s/w side as well as any work pertaining to embedded control systems.

If you want to end up with a software development job in the embedded systems field, make sure than you are proficient with linux as the demand for professionals with embedded linux seems to be increasing. If USU or Univ of Utah is offering it as an elective then take it, else just learn it on your own. It it manageable and should widen your job prospects.

NOT very active on edulix these days. For any questions or help, PM me (expect delayed responses).
Please do NOT tag me for profile evaluations.
03-23-2009 01:03 PM
Find all posts by this user Like Post Quote this message in a reply
amg_crce Offline
Moderator Emeritus
*******

amg_crce Offline
Moderator Emeritus
*******


Posts: 1,652
Likes Given: 0
Likes Received: 1 in 1 posts
Joined: Feb 2007
Unisearch: Link
Reputation: 18
Post: #5
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
CS Robotics may or may not have much much to do with embedded systems. I would imagine robotics is more of computer vision, AI kind of stuff rather than what you would look for in an Embedded Systems course. You should try and make sure what exactly goes on there.
03-23-2009 07:50 PM
Find all posts by this user Like Post Quote this message in a reply
PeacefulWarrior Offline
Edulix Senior Member
******

PeacefulWarrior Offline
Edulix Senior Member
******


Posts: 352
Likes Given: 0
Likes Received: 1 in 1 posts
Joined: Oct 2007
Unisearch: Link
Reputation: 7
Post: #6
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
(03-23-2009 07:50 PM)amg_crce Wrote:  CS Robotics may or may not have much much to do with embedded systems. I would imagine robotics is more of computer vision, AI kind of stuff rather than what you would look for in an Embedded Systems course. You should try and make sure what exactly goes on there.

amg_crce is right. Robotics here at the U is more focussed on Mechanics and Computer Vision. Infact, CS dept at U offers a Robotics track but all the courses and research is done by professors in Mech dept. So make sure you know, what is going on in a particular univ before making a decision.

@vivek,
nice initiative.

Although I have reduced my engagement on Edulix, I check it periodically. The best way to grab my attention is to PM me. I will reply if I have anything valuable to add. I am not the best person for SOP and university evaluations, so please don't waste your and my time PMing me about it.
(This post was last modified: 03-24-2009 09:26 AM by PeacefulWarrior.)
03-24-2009 09:26 AM
Find all posts by this user Like Post Quote this message in a reply
Zodiac Away
Edulix MVP
*********
Edulix MVP

Zodiac Away
Edulix MVP
*********
Edulix MVP


Posts: 605
Likes Given: 1
Likes Received: 10 in 8 posts
Joined: Jun 2008
Unisearch: Link
Reputation: 39
Post: #7
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
Ok I thought I will share whatever info I have gathered over the past few moths about how embedded is at various universities.

Let me start first with Arizona State University.

Going by the recent posts from my seniors, girish and Rooju, on the university specific thread on ASU, embedded is not that great at ASU. (refer post number 1354 and 1360).

Most of the research is under an assistant professor called Aviral Srivastava. He is doing some research work on microarchitecture, compiler optimization and stuff.

Apart from this, ASU is one of the few universities that offers a course dedicated to Embedded Linux (kernel modifications, device drivers, learning about designing, testing and then evaluation). On the downside, prof Miller who handles this course is not currently involved in any major research work in the operating systems side.

Although the research work may not be all that great, but I think if you are considering job opportunities on the embedded side, in today's market there is a huge demand for professionals who have expertise is device drivers, linux kernel programming and related stuff... atleast that is the trend that i have observed having been in this field for the best part of last two years.

Embedded systems falls under the CS dept purview.

For people interested in hardware software co design, SoC design and reconfigurable computing like myself, Prof Karamvir Chatta is doing some active research work (decent) in these fields.

Overall, embedded is not that bad at ASU. It is not brilliant as mentioned in some threads but it is a decent option to consider.

NOT very active on edulix these days. For any questions or help, PM me (expect delayed responses).
Please do NOT tag me for profile evaluations.
03-24-2009 04:30 PM
Find all posts by this user Like Post Quote this message in a reply
mastanmangu Offline
Edulix Frequent User
**

mastanmangu Offline
Edulix Frequent User
**


Posts: 37
Likes Given: 0
Likes Received: 0 in 0 posts
Joined: Oct 2008
Unisearch: Link
Reputation: 0
Post: #8
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
gr8 initiative tulsi Applause
well like the vlsi thread i would like to know the research areas,the kind of work to be done there, job prospects and even univ's giving admit with a perspective of an Electronics guy like me.
i am under hell confusion to opt for embedded or vlsi (mostly digital) becoz of the reason that i love programming but want to be in electronics related field. current sem course Microprocessor(6th sem) has shifted my interest towards embedded so unable to decide wat to do.
one more doubt
does one needs to have command over languages like java , python etc or having a knowledge of c/ c++ is sufficient
thanks in advance 4 reply

(This post was last modified: 03-25-2009 03:31 PM by mastanmangu.)
03-25-2009 03:14 PM
Find all posts by this user Like Post Quote this message in a reply
Zodiac Away
Edulix MVP
*********
Edulix MVP

Zodiac Away
Edulix MVP
*********
Edulix MVP


Posts: 605
Likes Given: 1
Likes Received: 10 in 8 posts
Joined: Jun 2008
Unisearch: Link
Reputation: 39
Post: #9
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
(03-25-2009 03:14 PM)mastanmangu Wrote:  well like the vlsi thread i would like to know the research areas,the kind of work to be done there, job prospects and even univ's giving admit with a perspective of an Electronics guy like me.
i am working on the draft content. I will post it here as soon as it is ready.

(03-25-2009 03:14 PM)mastanmangu Wrote:  i am under hell confusion to opt for embedded or vlsi (mostly digital) becoz of the reason that i love programming but want to be in electronics related field. current sem course Microprocessor(6th sem) has shifted my interest towards embedded so unable to decide wat to do.
one more doubt
does one needs to have command over languages like java , python etc or having a knowledge of c/ c++ is sufficient
I can understand. I have been in your position and I was as confused as you are right now. The good thing about the embedded domain is that a good knowledge of hardware and software is essential.
So I would suggest you to pick up courses carefully so that you get the best of both the h/w and s/w worlds.

Strictly speaking about skill-set for a person working on the embedded systems- s/w side, I would suggest the following (this is as per the suggestions by embedded systems s/w developers who have spent close to 10 years in this domain) :
1. C: Be strong in C. This is a must.
2. C++ : As processor speed increases, along with increase in memory density, application development will have to move towards OOP for maintaining huge code bases.
3. Scripting language: Testing in embedded systems involves a lot of actions that have to repeated over and over again. Learning a scripting language can help. PERL is a good choice. I would personally suggest Python simply because it is lot more better/easier than PERL. Even otherwise, knowledge of scripting can help in building good prototypes.
4. Linux: is becoming an increasingly popular choice of OS in embedded systems. This is not surprising because portability is not an issue with Linux, no royalties, access to source code, access to freewares etc. Learning is as simple as getting any popular Linux distro on your PC and then just simply playing around and trying new stuff. I would recommend Ubuntu.
5. GUI tools: with embedded handled devices becoming more and more fancy, knowledge of any ui designing utilities could be a big plus Qt on linux is a nice choice. Knowledge of Gtk is not bad either.

There! Hope that helps!

NOT very active on edulix these days. For any questions or help, PM me (expect delayed responses).
Please do NOT tag me for profile evaluations.
03-25-2009 04:57 PM
Find all posts by this user Like Post Quote this message in a reply
Zodiac Away
Edulix MVP
*********
Edulix MVP

Zodiac Away
Edulix MVP
*********
Edulix MVP


Posts: 605
Likes Given: 1
Likes Received: 10 in 8 posts
Joined: Jun 2008
Unisearch: Link
Reputation: 39
Post: #10
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
Continuing...
In this post, ill cover Delft University of Technology aka TU Delft.

Delft is one of the top ranked schools in Netherlands and very well known throughout Europe. As per the trend that I have observed in their decisions, they give quite a lot of importance to your undergrad scores.

Delft offers a MSc program in Embedded systems as a part of the 3TU. The program is heavy on course work but there is a lot of emphasis on research work. I felt that the Embedded systems program is more focused on the software side and the MSc program in Computer Engineering gives good exposure to both the h/w and s/w side equally. You can apply for for either one depending on convenience.

Talking about research work, they seem to be showing special interest towards reconfigurable computing and micro-architecture. I have also seen other projects on compiler design and optimization. In my view, the research work is pretty strong here. Maybe my senior rider_of_rohan can shed more light on this.

There are quite a few opportunities to get internships around this place. The popular belief is that for job opportunities, companies based around this area prefer the Dutch-speaking locals over foreigners.

You should consider Delft if you want to do your masters in Europe and want to specialize in Embedded Systems.

NOT very active on edulix these days. For any questions or help, PM me (expect delayed responses).
Please do NOT tag me for profile evaluations.
03-26-2009 09:29 PM
Find all posts by this user Like Post Quote this message in a reply
sriramg1987 Offline
New Edulixian
*

sriramg1987 Offline
New Edulixian
*


Posts: 9
Likes Given: 0
Likes Received: 0 in 0 posts
Joined: Nov 2008
Unisearch: Link
Reputation: 0
Post: #11
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
hey vivek...thanks so much for the info.I wud like to ask u if USC is good for embedded systems...although expensive,the univ is damn gud but i am not able to figure out if it is good for embedded...kindly shed light on this matter...thanks..
03-26-2009 09:41 PM
Find all posts by this user Like Post Quote this message in a reply
Zodiac Away
Edulix MVP
*********
Edulix MVP

Zodiac Away
Edulix MVP
*********
Edulix MVP


Posts: 605
Likes Given: 1
Likes Received: 10 in 8 posts
Joined: Jun 2008
Unisearch: Link
Reputation: 39
Post: #12
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
@sriram
From what I've seen, I dont think USC offers anything specific on embedded systems. I understand that it is a good university and it is sort of an 'any person's admit' university. But no, there is not a lot of work going on with respect to embedded systems here.

NOT very active on edulix these days. For any questions or help, PM me (expect delayed responses).
Please do NOT tag me for profile evaluations.
03-26-2009 09:52 PM
Find all posts by this user Like Post Quote this message in a reply
sriramg1987 Offline
New Edulixian
*

sriramg1987 Offline
New Edulixian
*


Posts: 9
Likes Given: 0
Likes Received: 0 in 0 posts
Joined: Nov 2008
Unisearch: Link
Reputation: 0
Post: #13
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
hey thanks a lot vivek...but i m kinda confused...between UT arlington,uncc n usc...uncc is good for embedded..n i want to focus on the hardware part of embedded...but usc has a brand name and is a top univ..which univ should i select in such a scenario..???
03-26-2009 10:56 PM
Find all posts by this user Like Post Quote this message in a reply
Zodiac Away
Edulix MVP
*********
Edulix MVP

Zodiac Away
Edulix MVP
*********
Edulix MVP


Posts: 605
Likes Given: 1
Likes Received: 10 in 8 posts
Joined: Jun 2008
Unisearch: Link
Reputation: 39
Post: #14
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
@sriram
It seems like you are very intent on USC. Smile
Since you are interested in the hardware side, why not major in digital VLSI. I think in that case you can pick up the Msc program in computer engineering. They also seem to be doing some research work in parallel processing and computer architecture but I am not sure about the quality of this research work to be honest. Maybe some seniors from USC can help you out on this.

You can also target the Msc program in Electrical engg (VLSI design specialization) but I think it is too focused on analog VLSI, core filter design concepts and stuff.

UNCC is a safe university in my opinion. Embedded is quite good but not so sure about the current research work @ UNCC. In my opinion, I doubt if it will be that great. Looking at the program, it seemed like it was oriented towards the hardware side.

At the end of the day, I would suggest you to pick the university where you will be majoring in something that you like doing rather than going to a top university and majoring in something that you don't like.

Think about it. The choice is ultimately yours.

NOT very active on edulix these days. For any questions or help, PM me (expect delayed responses).
Please do NOT tag me for profile evaluations.
03-26-2009 11:46 PM
Find all posts by this user Like Post Quote this message in a reply
sriramg1987 Offline
New Edulixian
*

sriramg1987 Offline
New Edulixian
*


Posts: 9
Likes Given: 0
Likes Received: 0 in 0 posts
Joined: Nov 2008
Unisearch: Link
Reputation: 0
Post: #15
RE: Embedded Systems - Doubts, Career Prospects, Universities
0
0
heyy thanks a lot for ur opinion vivek....i have got a much clearer idea now...
03-27-2009 12:44 AM
Find all posts by this user Like Post Quote this message in a reply
« Next Oldest | Next Newest »
Post Reply  Post Thread 


View a Printable Version
Send this Thread to a Friend
Subscribe to this thread