State of Analogue & MS Design today?

[bernie21]

Hi,

I was reading an article/report (http://www.designers-guide.org/Design/tdd-principles.pdf) about how Analogue and mixed-signal design methods are too bottom-up when they should be top-down (like the modern digital methods). Browsing around, I haven't found much material published on this topic since... What tools exist to fix it?

I've found Spectre and HSSPICE as possibilities, but I don't know to what extent these address RTL verification and synthesis. Also, are there automated methods to verify "gate-level" analogue circuits?

Thanks in advance!

 

[jim hardy]

in my humble opinion

small efficient organizations can do fine with bottom up methology.

large bureaucratic ones must resort to that author's methods

and that's why so much of large corporations' work is farmed out to smaller ones

Parkinson's Laws should be required for both engineering and business curricula.

"A methodology is a method that went to college" - Mark Twain

 

[bernie21]

in my humble opinion

small efficient organizations can do fine with bottom up methology.

large bureaucratic ones must resort to that author's methods

and that's why so much of large corporations' work is farmed out to smaller ones

Parkinson's Laws should be required for both engineering and business curricula.

"A methodology is a method that went to college" - Mark Twain

Thanks for the quick response Jim.

I agree with you on small organizations since I doubt they would pursue large, complex designs right out the gate. As for large corporations, I doubt they would outsource much of their work to smaller corporations - it would be costly and difficult to line up efforts for multiple blocks... Do they?

For high-tech, I expect both engineering and business curricula do employ Parkinson's Law; especially in ASIC design. I'm just not positive.