What is the future for CMOS RFIC Design Engineers?

cmos rf ic future

Many companies doing RF CMOS have failed recently, espcially in Wireless LAN.
Companies like Broadcom, Atheros, Marvell and TI say they have RF IC's in CMOS but where is the proof? Where are the products?

IS RF CMOS just an academic thing? Is their a future in RF IC Design in CMOS in industry or just wishful thinking? Noise is too high, power consumption is too high, mask costs for 0.13um and 90nm RFCMOS are way too high compared to 0.18um RFCMOS so cost vs. bipolar is not an issue any more.

What is the future? Comments and Discussion welcome.

The future is in CMOS!
If it's true that some companies are having troubles with pure CMOS, it's also true that the RF is always driven by baseband capabilities.
Now we are in a phase where BiCMOS (mostly 0.35,0.25 and sige) is the must if you want product today. I am designing using it.
The future (the open question is exactly when) is in the integration of projects of RF & BB, the company who has the best cooperation with analog and digital is the winner. The integration is in that way, with some step towards a single IC, through modules, SIP and so on.
The problem is that the analog part is tachnologically some step behind the digital part.
Costs are driven only by volumes: some year ago a 0.25 SiGeBiCMOS was very high cost, today 65nm CMOS is the highest cost, while 0.25 SiGeBiCMOS is relatively cheap.
When someone will be able to have a product with RF & BB in one die, he'll be the winner. And the possible choice is only CMOS.
Mazz

i don't think a single chip radio or any transceiver is the answer for anything. sure it might be cool and a lot of professors and universities see this as the ultimate but it does not make sense from a cost and performance point of view.

what we have now, the system in a package (SiP) vs. system on a chip will win for mixed analog/rf/digital applications, thus rendering the usefullness of cmos gone.

the cmos device is not limiting the performance the process is (and i think always will) limit the performance.

just my opinions.

I see.

Consider this: the evolution from discrete design (different technologies, large area) to RF IC has been driven by cost, not by performances.
The evolution from classic superhet to low IF/zero IF has been driven by BB capabilities and by cost.
Cost is always present.
Do you think that SIP is cheaper than SOC?
I think that the only open point is when not what.

To come back to the original question, what an RF IC designer can do today is to integrate in his own package of knowledge the more he can on communication & digital part of the system he's working on, consolidate relationships with BB and system designers to win the race.

Also just my opinions.

Mazz

Point you are trying to make is that "RF IC" Designer is no longer just an RF/Analog/Baseband IC Designer, but really a Radio System IC Designer. A designer of compex radio systems.

I think UWB is one of the future you can attend. Or DTV or DMB application maybe are another choices.

Yibin.

Actually, there are some papers about CMOS RF chips from industry in recent ISSCC. It's definitely an attractive way to do CMOS RF stuffs since the cost will be substantially reduced. This is the dominant factor for commercial products.

Intel hasn't been sucessful in cmos rfic.

Papers don't mean anything, I am talking about products!

Atheros, Marvell, Realtek have all successfully produced CMOS RFIC for WLAN. The products are out there!

Puppet1 said:

Intel hasn't been sucessful in cmos rfic.

Papers don't mean anything, I am talking about products!

Click to expand...

ISSCC papers are not so far away from products

For ieee papers, there is a must to produce some produce or at least prototype to show some evaluation results right?

From opinions above, it is seems that for RF engineer today, the most important thing is to combine all the circuit blocks, strong in circuit and also with some knowledge background of communication?

i think the cmos rf is the trend of soc .

I think UWb will be a hot area .