Abstract
The initial digital benchmarking survey can be segmented into four distinct
sections. The first section, which includes questions 1 and 2 are designed to
determine the respondent' s background and primary job function. In order to
insure the validity of the survey, it is important to verify that the
respondents are qualified to answer the questions. It was also designed to
ensure that there was a representative mix of system makers, power converter
manufacturers and semiconductor manufacturers. In addition, the last part of
the survey (immediately after question 12) asks for respondent information,
such as company and location. This information was used to verify that
companies or alliances between companies could not skew the results.
The second part of the survey includes questions 3 through 5, these are
designed to rank digital power IC companies. Questions 6 though 8 make up the
third part of the survey and are designed to rank digital power converter
companies. The final part of the survey, questions 9 through 12, is designed
to rank the overall position of both digital power converter and digital power
IC companies.
The structure of this questionnaire lends itself to a straightforward,
quantitative analysis. To determine top-level mind share, each respondent was
asked to list their top three companies for each category in order. A company
that was listed first received 3 points, second received 2 points and third
received 1 point. These points were totaled across surveys to determine
leaders.
To determine rank on individual aspects of digital power, the respondents
answered two types of questions. The first was a ranking question with
features/functions such as Price, Interoperability, Design Tools, Pure Digital
Solution and Hybrid Analog/Digital Solution ranked 1 through 5, with 5 being
most important. In order for the response to be valid, the participating
respondent could not use any number more than once, and could not leave any
attribute unranked. Then for each company mentioned, the respondent rated the
company on a scale of 0 (poor) through 10 (excellent) based on how the company
performs on that feature/function. The rank and rating were then multiplied to
determine how that respondent felt about that company. For example, if a
respondent said "Company A" was a 10 on Price and rated Price as the most
important (5), the company got a score of fifty (10 x 5 = 50) for the Price.
On the other hand, if Interoperability was ranked a 3 and the company rated a
6, the company received a score of eighteen (3 x 6 = 18). Companies were then
compared based on their average scores across respondents. Prior to conducting
this analysis, all "own responses" were ignored. This means that any time a
representative from "Company A" mentioned "Company A" that response was
removed from the data set. This was done to stop any company from "stuffing
the ballot box." The same procedure was checked for alliance/standard
membership. However, since alliance/standard membership did not have a
significant affect on the findings, alliance membership was not adjusted for
in the final results.
The second follow-up Digital IC survey consists of 5 questions. Questions 1
and 2 are designed to determine what level of integration, and which features
and functions should be included in a digital IC. Question 3 is intended to
determine what type of communications protocol(s) and bus(s) should the IC be
able to use. Question 4 identifies the target markets for the IC
manufacturers. The final question, number 5, is based on the markets checked
in question 4. In this question, the respondent ranks the desired features and
functions from 1 to 5 in order of importance (5 being most important and 1
being least important). In this question, respondents were allowed to use a
number more than once. In order to thoroughly evaluate the responses in this
section, a number of comparisons and references will be made to Darnell' s
earlier 2007 study "Digital Power Electronics: IC Product Introduction Trends".
The final survey was a short two question survey designed to determine what
percentage of the participating companies focused on standard designs vs.
custom designs. Unlike the other surveys, this questionnaire allowed the
participant to add their own personal opinions and comments on the percentage
breakdown of custom vs. standard design in the area of digital power
management.
| 
View Cart
