Planning to change domain from SC to PD, whats your openion frineds?

Hi Friends,

Its exactly 6 years today as i started my career in VLSI field!!!
Started working with "SoC Subsysterm Design and Integration" moved to "IP verification" then worked on "RTL coding" of few Adapter IPs. Then again worked on RTL integration. All this happened in 2 years of time frame. Then due to recession I moved out of Frontend Domain to Backend where I got opportunity to work on Memory Compilers. It sounds bit messy initial days to understand what is GDS/LEF/ANTENNA/SI but it was very interesting and i could easily cope up with that in 3 months time. After working in Memory Compilers for 1 year, I got an opportunity to work in Standard Cell. Here from last 3 years I am working in Standard Cell library design -Schematic/layout design, Characterization, different EDA views generation like celtic/voltagestorm/verilog modelling etc...

Now I am realizing following drawbacks in current Standard Cell Library Design:
1> Work is more repeatative in nature.
2> Schematics we draw are for only small circuits, its nothing when compared to IO domain or Analog domain.
3> Layout work is very tedious in nature, sounds interesting when drawing first time but most of the time even a diploma guy can also do that
4> Characterization is very interesting but most of the time we spend time in firing jobs/monitoring jobs due to multiple PVT corners in package.
5> Major thing is the JOB opportunities. There are only countable companies who work in this domain and its very risky if one needs to hop a company in near future.


So, I am planning to move to Physical Domain since I have confidence that I worked on Standard Cells/Memories/RTL coding/SoC Subsysterm Integration/Verification, I myself realizing that I can do very well if I move to Physical Design - RTL to GDS2. Here I planning to work deeply in Timing Closure and Power Grid Analysis. I thought of moving to IO/Analog (ADC/DAC/PLL design) domain, but due to less confidence on CMOS circuit theory I don't want to take chance in IO/Analog domain.
As there are more job opportunities in Physical Design and also more chances of setting up a business in Physical Design, I am working towards learning Physical Design related stuff.

I hereby request you friends that is this move which I have planned is correct or not ? Please help me in deciding this. Your comments are most welcome!!!

Thanks

I have been working in PD area for the last 5+ years. Most of the points you mentioned above are also true for PD especially the statement that even a diploma guy can do that..

Hi Matter, Could you please elaborate more on your work? I assume as you gain more experience in Physical Design specially in Timing Closure/Power verification, you get more knowledge on System Architecture and with this there are more challenging opportunities to work as System Architect. Be it a Job Hopping or starting a new business, PD is more safe compared to Standard Cell. Please elaborate on negative side of PD.. Thanks

I feel PD as the very good domain to work on. It is the most challenging work to do now especially with limited time to market these days.

Well PD, involves sound knowledge of CMOS circuits... nad will give you an exposure to work/concentrate on other implementation challenges such as synthesis,prelayout sta,post layout(ofcourse after PD). Thus, i feel the role is indeed challenging. As far as the negatives sides are concerned. Sometimes it can make your weekends go Hell... thats what i feel...

cheers

No, PD is very much separated from system architecture. In fact in I have gained very less knowledge about out system architecture through PD. I believe that what you mentioned will be true for LD (logic design) or test engineers.

Rather than elaborating on the negatives, let me tell you this. Have you heard about the term 'Polygon pushers' ? thats what PD guys are used to be called as. Search for this term in google and you'll see what I mean.

Btw, I am not trying to say that PD is less challenging. I have also spent endless hours during tapeouts. My point is that is that both std cell design and PD are on the same level as far as the job itself is concerned. In both cases, your work will be highly scripted or automated. In both cases, you are not going to invent anything new. But from a # of job opportunity perspective obviously # of PD will be higher than # of std cell jobs. I am not sure about payscale.

Sidetrack a little.
For an experienced FE engineer (11 yrs) in wireless communication arena, what's the next field to go where it's logical and can reuse some of the knowledge you acquired in the FE? In my country the wireless & ASIC industry is dwindling and very unstable, and only 3-4 players remaining as of now. What are the fields which is more stable and can use some of the skills I knew in FE?
I am basically an all-rounder besides FE work, I also knows most of the digital/sw aspect of popular wireless chips like WLAN/BT, but am feeling the glass ceiling now and idling of my skills.

Thanks for your input.

Hi kel8157,

All i would say is "Don't worry". Specially when you have very good understanding on wireless chips!
I too gone under same situation when working in FE domain 4 years back. But believe me, you can do very well with the experience you have. Couple of my colleagues who are also in the same domain performing well, they are members in IEEE mobile standards,also they published few books, they are working now in USA. BTW which country you are talking about? I believe with your experience level you can perform very well in Manager/Architect/Consultant roles.