what is hot-topic in digital design now?

[maitrey17]

Hey,

I am a new member on board. I ma a Grad student in Digital Design n working on project for ASIC Design.

In reply to your Question,

The QUestion is really a broad minded. In Digital Design if you are talking about FPGA's and ASIC's then,

Die - yield.

Signal Integrity and Crosstalk.

Heuristic Algorithms that work behind the EDA Tools and give you the "supposedly" optimum Placement- Partitioning- Routing - Floorplanning Solution is also a big deal these days. Because not one particular Algorithm works and any kind of problem. So this makes Physical Layout harder and harder.

Above all this, I believe most imp. topic right now is "where is the end of the ever increasing process technology?" Right now we are moving towards 65nm technology where , NONE of the Industry EDA TOOLS have the capability to become compatible with 65nm technology where we are moving. All EDA tools have been built only uptill 90nm technology. So, when we move our forces for 65nm, we already have crosstalk and signal integrity issues in 90nm so think about it in 65 nm. Also, the yield is very very less at preset....so think what will it be at 65nm.....

I believe that this is the biggest Revolution as far as every single thing is concerned. all eda tools have to be re designed and all algorithms have to be re considered. in this process, there are few leading companies that are having a hard time with this all issues.

 

[Puppet1]

i have seen work and met people doing genetic algorithms for EDA design.

do you personally think we can get over the barrier to 65nm in EDA or will it take a long time and possibly never?

 

[wkong_zhu]

In tools, soc verification, and physical signal integrity tools is bright.

In PC bus USB OTG, PCI Express is hot.
wireless USB is being built now, must be it

 

[maitrey17]

i have seen work and met people doing genetic algorithms for EDA design.

do you personally think we can get over the barrier to 65nm in EDA or will it take a long time and possibly never?

you are right! Simulated Annealing and GA are really powerful Algorithms. Recently, in my class we had many people present discussion on these topics.

As far as 65nm, I dont have much experience so dont quote on my word but yes, We have to cross the barrier may be in a bit longer time. but this is definately going to give us lots of conclusions.

I mean, even if we will be able to go for 65nm, ppl wont prefer to go in that kind of complexity because of higher NRE and work involved. Other companies are going to get your market if you are late.

But then, people will set the boundaries of how far they can go as far as the process technology is concerned. They will use 65nm if they have time and resources to devote...but if they wanna get the work done and in quicker turn around time then definately 65 is not the answer.!!

Let me know your views. By the way, what is your profile? what have you worked on and let me know something about you.

 

[feel_on_on]

FPGA and System on chip!

 

[ldhung]

Convert designs from FPGA to ASIC is the hottest topic to me

 

[icon]

as someone said verification is the hot one.about 70 % of product time is spent for verification alone..and now new verification technologies are also emerging..

 

[the_penetrator]

I would suggest the following:

1. Heterogeneous SoC
2. Biometric chips
3. Physical processes emulation engines (watch out for the Physics Processing Unit...)
4. High-level verification
5. ***Rapid*** Software tools generation (to support processor design)

cheers

the_penetrator©

 

[chinara]

I believe the channel decode is another high profit field.

 

[goalied00d]

I'd say console gaming hardware,

 

[Puppet1]

is asic and fpga design dead?

 

[power-twq]

I think the most hot topic in IC design is low power, high performane,

low cost.

Can anybody share any idea? I think SoC maybe one.

 

[Puppet1]

how about communication algorithms in fpga/asic

 

[dinnu_don]

HI
i can have exp on the specman verify the SoC and can find the lost of bugs which could be removed in the design stages rather leting it to go in silicon

The feature of the randomisation which is really produce the scene which human ignore in the bigger end

 

[zhaoyimiao]

Uses digital circuits to achieve the RF module's function is a very interesting area.

Added after 1 minutes:

Digital design for still image processing like digital camera is very hot I think.

 

[silverbyte]

I think it is DFM - Design for Manufacturing.
At sub 90 nm levels, chipd yields are going lower and lower. I think DFM is the next hot field to come up

 

[jjww110]

shorten verification

 

[justin999]

low power
dft
dfm

 

[eeeraghu]

PCI Express, AMBA, OCP/Ip SOC's regarding the protocol level,
Verification With E or Vera,
Fututre might be network on chip.

 

[dak-ju]

Hi,

I think the hot topics in digital design are

o. DSP based audio/video processing
o. High Definition Television
o. Low power mobile processors

Regards
dak-ju