Who has done all that A/D-converter research?


ADC publication ‘power law’ curve: A handful of actors produce a lot, while the majority of organizations produce an accumulated total of one or two papers. Only papers reporting measurement results were included.

Of the approximately 1500 scientific papers covered in the ADC survey reported in [1]-[2] – also mentioned here, here, and used for the underlying data set here – who was actually responsible for all that work? Well … I could give you a complete list of author names – a very long list of names – but I’ll settle for a summary instead:

According to my data, there are about 270 different organizations that have at any time in history implemented, measured, and scientifically reported at least one A/D-converter IC. Now, that’s an average of five and a half papers per affiliation during the 36 years of monolithic A/D-converter integration [1].

But life isn’t that fair. In fact, there is no such thing as a representative average in this field. Instead, the publication count per organization roughly follows a power law curve as seen in the bar plot above. A few organizations (20–30) completely dominate the field, while a large number of organizations are represented by only one or two contributions.

I want to emphasize that there is absolutely nothing wrong with organizations that belong to the latter category. It’s fully possible (even likely) that these organizations deliver research results of the highest scientific quality – just that their focus is not on physical implementation. When looking at companies, one must also consider the possibility that a lot of activities may be going on “under the lid” for commercial reasons. But when it comes to publicly available research results, a relatively small group of actors have produced the majority of results, and among the top 30 there is an even distribution between companies and academic institutions. The top three affiliations are companies, and the following five bars represent the five most productive universities.

Note also that affiliation was determined from the first author of each paper and that work done at one affiliation is sometimes reported under the author’s current affiliation. I’ve tried to assess the true origin of the work as best I’ve could, e.g., by checking for affiliation footnotes showing past and present location. Since there is no way to know to what extent such information was always included when appropriate in all papers ever published since 1974, one must allow for some error margin here. And that’s one good reason to not start naming affiliations in this post.

Academic vs. corporate

ADC paper distribution between companies, universities, and “institutes”. Affiliation is determined by first author.

Instead we’ll check how the accumulated amount of ADC research papers divides between corporate and academic research? Not being really sure how to categorize some affiliations, I’ve introduced a third category, “Research Institutes”, for organizations that aren’t universities, but not classic companies either. In this category I have placed IMEC, ETRI, the Fraunhofer-Institute, ITRI, and the Space Research Institute. As illustrated by the pie chart, the total amount of scientific ADC papers published since 1974 divides into Academic (49%), Corporate (48%), and Institute (3%) when grouped by first author, and as much as 63% of the papers have at least one author affiliated with a company. A very strong impact from the corporate world, indeed.

Geographic region

Accumulated amount of ADC research papers by geographic region. Affiliation determined by first author.

Finally, we can look at the regional affiliation by continent: The accumulated total of ADC papers divide between North America (50%), Europe (~27%), Asia (~23%), and Africa (0.1%).

Africa is represented in my data by a single paper [3]. In comparison with North America, Europe, and Asia, a single paper may seem a bit slim. On the other hand it can be noted that, according to my data, Africa outperforms both South America and Australia, and that was not an obvious result – at least not for me.

Disclaimer: It is fully possible that I’ve missed some publications from these three continents, so please post references to them in your comments here if you know of any. Just remember that what I’ve counted here are papers about measured ADC implementations where the measurement results are reported in the actual paper or accompanying slides, and the continent scores according to the first author.

Now, didn’t this post make you at least a little bit curious? Don’t you start to ask yourself:

  • Which are those top-3/5/10 affiliations?
  • Which companies have given up on ADC research, and who are the new players that recently joined the game?
  • What are the trends among companies/universities in North America/Europe/Asia?
  • Is the whole world going in the same direction, or can we see any regional differences?
  • What’s the average publication rate among the top five companies, and how does that compare to the top five universities?
  • Are any of the top five slowing down or perhaps picking up an unusually strong wind?
  • What’s happening to the pie slices – which ones are growing, and which ones are shrinking?
  • Etc, etc …

Well … my boss (that’s me, actually) tells me I’ll need to save some information for ADMS Design AB and its clients too, but I will tell you something interesting in the next post nevertheless. You probably don’t want to miss that one if ADC research and development is your thing. So stay tuned …

Don’t forget: You can use the ‘Subscribe’ function to get an e-mail every time there’s a new post.

Edit: As mentioned here, this post is now cited in a scientific paper by Fuiano et al. [4]. Should you wish to do the same, you can cite it as suggested below [5].

See also …

ADC Survey Data

References

[1] B. E. Jonsson, “A survey of A/D-converter performance evolution”, accepted for presentation, IEEE Int. Conf. Electronics Circ. Syst. (ICECS), Athens, Greece, Dec., 2010. [@IEEE Xplore]

[2] B. E. Jonsson, “On CMOS Scaling and A/D-Converter Performance”, accepted for presentation, NORCHIP, Tampere, Finland, Nov., 2010 [@ IEEE Xplore]

[3] E. Hegazi, and N. Klemmer, “Accurate Modeling of Noise in Switched-C Analog-to-Digital Converters”, IEEE Trans. Circuit and Systems, Pt. I, Vol. 52, pp. 2319-2326, Nov., 2005.

[4] F. Fuiano, L. Cagnazzo, and P. Carbone, “Data Converters: an Empirical Research on the Correlation between Scientific Literature and Patenting Activity,” Proc. of Int. Workshop on ADC Modelling, Testing and Data Converter Analysis and Design (IWADC), Orvieto, Italy, June, 2011.

[5] B. E. Jonsson, “Who has done all that A/D-converter research?,” Converter Passion, Nov. 6, 2010, Available: http://converterpassion.wordpress.com

About these ads

4 responses to “Who has done all that A/D-converter research?

  1. Pingback: Is A/D-converter research in Europe about to die? | Converter Passion

  2. The number of publications might depend on the fact that companies in Europe focusing on A/D converters have diminished. Most companies active in the A/D semiconductor field are of US origin and their interest in sponsoring research in Europe is limited.

    Exceptions from this rule exists primarily in a few countries (Germany, France, Netherlands primarily) where semiconductor companies developing A/D building blocks for use in advanced mixed signal circuits are still active. The academic interest is linked to industrial activities and it is therefore logical that also publications from academia in Europe are becoming more rare.

    It is however interesting to note that more and more of the mixed signal development is taking place in the digital domain including linearisation, error correction etc. Such research is not necessarily labelled A/D research so maybe there is more of research going on in Europe than what is visible in the survey?

  3. @Gunnar:
    Thank you for sharing your views and insight on this matter.

    As you say, there is a correlation between industry and academia in that research grants are often given based on industrial relevance for the country or region in question.

    The survey includes every single paper I could find that reported a characterized A/D-converter IC implementation (mainly in these sources). Whether the main effort was in the analog design or in digital linearization did not matter for inclusion. But a criteria for inclusion in the survey was that a chip was actually manufactured and measured, and that the measured performance was reported in the paper.

    With a purely digital approach it becomes less important to implement your own ADC for experimental purposes. If the main interest is to develop better linearization techniques, you might jus as well use the output data of an existing design. Such papers were typically not included in the survey, and therefore the extent of such research is invisible in the study.

    I’m aware of several such activities in Europe alone, and a surprisingly large number of small and medium sized companies that are actually doing ADC design too. In addition to “the big ones” that is. So I’m not as worried for Europe as the headlines here might suggest.

  4. Pingback: Converter Passion Citation Boost | Converter Passion

Comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s