What is the latest development of Analog EDA?

I am a master student who want to go on my Ph.D. study on Analog EDA. Most research experiences of me are related with vlsi physical design, so are there some papers giving the general description of this field with emphasis on latest development and trend?

It will surely be necessary for a freshman like me in this field. Great thanks to you.

My wish list
1) D/A co-sim
2) Accuracy && speed
3) LPE

hoteagle said:

I am a master student who want to go on my Ph.D. study on Analog EDA. Most research experiences of me are related with vlsi physical design, so are there some papers giving the general description of this field with emphasis on latest development and trend?

It will surely be necessary for a freshman like me in this field. Great thanks to you.

Click to expand...

Hi ,
I ever heard a analog stnthesis tool by antrism .. but I don't know it work or not .. another new tool called "NeoCircuit" it is
hspice optimation tool , and automatic re_size analog circuit /layout ..

I think analog Rd need some analog tool for design , like

analog filter design --> from active_filter , SCF , Gm_C filter ...

automatic OPA design tool --> can synthesis a OPA , if we assign topology gain Bandwidth ... and process ..

I've heard about companies doing EDA for analog circuit synthesis - one of them is Analog Design Automation. I think their approach is based on developing some kind of an expert system plus optimization. Personally, I hope they won't succeed in the forseeable future or great part of the fun doing analog design will be gone. But this is just my humble oppinion

thanks for your message,
*****************************************
1. "tlihu": the first item you mentioned is really hot in my view. it relats with mixed-signal design and simulation which is also the main research area I am interested in (of course, EDA for this area). The second is common sense for all circuit design. The third, I think there are many good EDA tools and it relates not only with analog, but digital and physics. I have used Dracula from Cadence and EPIC(PowerMill, TimeMill) from Sysnopsys, but not familiar. Magma does well in this area for DSM design.
*****************************************
2. "andy2000a": how do you know these analog EDA tools? Do you often use analog tools? The most often eda tools I or my friend using is Virtuoso Mix-Signal design Enviroment. I don't know is there some company use analog synthesizer to do there analog design. Most common thing I know is to adopt some analog IPs and re-design them to make them suitable to the definited process and SOC chip enviroment. Am I right? I don't know much about the analog eda tools' usage condition in company.
**************************************
3, "sutapanaki"
I agree with you My girl friend is also an analog design engineer and she told me the work takes pains and only person like her who are persisting on a certain thing for a long term can do it well. I think maybe I can resolve it with some simple setting and let all arduous things for computer doing. Then we will have more time to take a walk or go to movie house.

Thanks for your help and waiting for more.

Happy new year.

Hi hoteagle,

I have an idea for a support tool. I have painful experience with mixed signal projects having about 50 analog blocks and about 20 power modes.

So normal signal operation of the blocks is verified by simulation. Also cooperation/coexistance/coimmunity is simulated.

But what hurts is that some blocks do not switch off complete. They leak some nA up to mA. In simulation they leak pA in off mode. The reason is they have "analog floating gates". That is some nodes are brought to high impedance in off mode. But the potential of that node is dependend on uncontrolled process condition or varies with time. Further this node drive a device which increase the current consumption..

To my understanding a tool which highlight such improper designs is not avaible and also not discussed in research (IEEE Circuit&Systems and similar). It could work by traversing the flat netlist and using the properties of an exsiting operating point analysis.