Over 60,000 people in over 100 countries read FPGA Journal.
Apparently, you’re one of them.
While we may be the perennial pundits of programmable logic, we don’t deceive ourselves into thinking that FPGAs are some sort of universal unification language. We do, however, think that programmable devices like FPGAs will be one of the most important technologies in electronic design moving forward. With the rapid evolution of the global high-tech economy, our position in the FPGA space gives us a fascinating lens through which we can observe and prognosticate about the pending progress of electronic engineering.
People use programmable logic for an ever-widening range of applications. As such, we notice the FPGA Journal audience expanding and morphing right along with the shockwave of pervasive programmability that is cascading throughout the electronics industry. Over the past couple of months, we’ve done a survey of our audience and learned some interesting things about where FPGAs are going, how the technology is being applied and deployed, and (most interesting to us) what you all want from a publication like FPGA Journal.
The first thing we’ve noticed is that FPGAs are going global. For a long time, the lion’s share of FPGA consumption was in the United States. If the makeup of our FPGA Journal readers is an indication, the US is still the single biggest geography, but now the other geographies are growing much faster. The Americas now account for about 40% of our readers, where they were 49% just two years ago. Europe has shown rapid relative growth, moving from about 30% to 34%. Asia/Oceania accounts for the remaining growth, jumping from 21% to 26%.
Overall, that suggests to us that Asia is growing faster than any other region in terms of FPGA interest, followed by Europe, with North America growing at a steady but more relaxed pace. Why? We think it’s because FPGAs are moving into new application areas (but we’ll talk more about that in a bit).
When more engineers get interested in using FPGAs, more engineers sign up for FPGA Journal. That’s exactly the trend we’ve noticed over the past few years as our rate of new e-mail signups per outbound advertising impression has increased significantly. We estimate that there are well over 100,000 professional engineers actively working with FPGAs. Why is there a percentage that doesn’t read FPGA Journal? We were wondering that too. (So we asked, but more on that later as well.)
This increase is directly related to the expanded use of FPGAs in many new markets and the “lowering of the bar” of barriers such as BOM cost contribution of FPGAs, density limitations, power consumption, performance, design tool capability, IP availability, and many more. As these limitations are reduced or removed one by one, the reasons for NOT using FPGAs in particular applications disappear, the production volume at which FPGAs makes sense goes up, and more design teams adopt FPGAs into their projects.
In addition to seeing FPGAs in more systems, we’re seeing a migration of FPGAs into more important roles. FPGAs first landed on the board as “glue logic” connecting components with incompatible interfaces. Now, in many cases, they’re the centerpiece of embedded systems handling applications processing, DSP acceleration, primary I/O like Ethernet, USB, and PCI, and integrating peripherals and custom hardware into a single chip.
This move of FPGAs into more important roles has changed the model for acquisition dramatically. For a “glue logic” FPGA, the decision to use a part (and which part to use) was made quickly, at the last minute, by an individual contributor engineer – often without much consultation on the decision process. An engineer would be finishing up a design, would run into a couple of chips that wouldn’t talk to each other, and would quickly pick a suitable FPGA to map protocols, change voltage levels, and do whatever else was necessary to get the system to work.
With the FPGA at the heart of the system, however, the decision is made much earlier in the design process, often at a higher level of management, and usually with more consultation and collaboration in the decision process. Our reader survey statistics show this trend as well – in the increased percentage of readers at higher levels of engineering management, for example. Over 10% of our readers now list themselves as “executive management” or “engineering management/supervision” at systems design companies (up from about 7% a couple of years ago). More managers and executives are educating themselves about the latest in FPGA technology because more executives and managers are finding themselves in positions where they’re making important decisions related to FPGAs.
While our education audience has grown, it has been radically outpaced by our commercial audience growth. Two years ago, education accounted for 32% of our audience. Today, while the raw number is larger, the percentage has dropped to about 10%, and the mix of faculty-to-student has shifted to a faculty majority. FPGAs have always been a great platform for student projects. Development kits and software can be had for a song, and an EE student can pop off a sophisticated system-on-chip design with almost no budget.
The mix of companies you work for has shifted as well. Over 50% of you work at either original equipment design/manufacturing companies or as consultants doing design work for those companies. Another 15% are doing chip design (and presumably using FPGAs for prototyping/verification), and 4% are developing licensable IP (and presumably using FPGAs for prototyping and verification as well). Another 10% list yourselves as “research.” Take off the 10% education, and the remaining 11% or so are marketing and advertising folks, industry analysts, sales and distribution employees, and people working at “insider” companies like FPGA and EDA vendors. Two years ago, the “insider” percentage was much higher. Our theory is that we’d saturated that audience pretty well years ago, and the subsequent growth in other areas has reduced the “insider” percentage.
The real proof in the pudding for expanded FPGA market, however, is in the “target market segment” area. As little as two years ago, our audience was still predominantly doing communications and networking design – the staple applications that have supported the FPGA industry’s growth through most of its life. Now, however, over 23% of the audience says their products target consumer applications – a segment that was all but closed to FPGAs until very recently. Defense and aerospace has also shown significant growth with almost 1/3 of the audience targeting that market with some of their FPGA-based products. Automotive transportation, display systems, industrial automation, medical equipment, PCs and peripherals, scientific equipment, storage and servers, test and measurement, wired and wireless communication, and data communications all have double-digit percentages of our readers selling products into them. (If you’re one of those arithmetic savants and you are about to scream that the total is over 100%, relax. Some people design and sell products into multiple markets, so the percentages total much more than 100%).
So – what do you people want from the world’s #1 FPGA-related publication (humility, perhaps? OK, we’ll work on that.) According to survey responses and comments, most of you come here for the feature articles and webcasts about once per week. A smaller number come almost daily for news stories and press releases. You told us you overwhelmingly prefer web-based articles and stories to print-based versions (63% rated web-based delivery “very useful” while only 30% rated print “very useful”). Interest in streaming audio/visual content such as webcasts is up sharply, and over 88% of you said e-mail newsletters were still your favorite way to get “push” updates of information. Interest in new delivery mechanisms like podcasts and blogs is still on the low side, with over 50% rating both media as “not very useful” or “completely useless.” We’ll continue to monitor the latest emerging distribution mechanisms and will try to always serve you your FPGA information on exactly the plate you want.
Forums were also popular, with almost 60% of you saying that they were “somewhat useful” or “very useful”. With that in mind, we’ll try to improve the usability of Journal Forums. We’ve had enormous problems with spam there, so we’ve had to put a “moderator review” process in place for new posts. As a result, if you make a post, it can sometimes take hours (or even a day or more) to show up. We’re working on technical solutions to the spam problem that might let us open up access more. Meanwhile, go ahead and post. We’ll approve your comments as fast as we can so your voice can be heard.
Finally, the most useful information for us was the wealth of thoughtful and insightful comments you all have added to the end of your survey responses. We asked what you would do if we made you “publisher for a week.” Hundreds of you submitted comments and suggestions. Over 75% of the comments were some version of “Don’t change anything, it’s very good already,” or “no changes – it’s much better than average,” or “It’s already the best, don’t do anything different.” To all those people – thank you very much! (you buncha’ suck-ups and “yes” men) You are all now officially fired as publishers. The remaining 25% of you are still in the game.
Of those comments, people seem to like the fact that we avoid annoying advertising styles like pop-ups, roadblocks, peel-back corners, etc. We have to run ads, of course (it pays the bills), but we want those ads to provide useful additional resources from our sponsors rather than detract from your reading experience. We’ll continue to hold the banner-ad geniuses at bay, and you can continue to support the ones that provide you valuable goods and services while reading our publications in relative peace.
As you may have noticed, we employ a different editorial and writing style than the rest of the trade press. We got a number of comments related to this – and those comments fell into two categories. First, on the issue of incorporating metaphor, simile, sarcasm, silliness, alliteration, and iambic pentameter in technical articles, the majority of you asked for even more of those devices. A smaller group, however, asked us to change our editorial style to more formal technical journalism. We got comments like:
“The one thing that is really disturbing is the writing style of the articles. Suppose there a topic that interests me, I know that reading it at FPGA Journal will be a pain because of all the exaggerations and ‘mumbo jumbo’ writing style. Take for example – “Under cover of darkness, our black-clad insertion team would rappel down the walls of the super-secret Xilinx fortress in the desert.” this is not a very technical writing style. I have little time to go through these journals and when I do find something interesting I usually skip half of the paragraphs because of these things. I have to say however that the topics themselves usually are interesting. Please just keep in mind that as you are a technical newspaper, your customers expects (sic) technical writing. I hope that you will take my advice because overall this newspaper is very good.”
This went straight to our little editor hearts, and we’ve subsequently decided to omit these elements from all future articles. OK, just kidding. While we’d like to make everyone happy, the vast majority of our audience is telling us that the writing style is one of the primary reasons they prefer our publications to those of our competitors. We can’t take that away without disappointing a large majority of our audience. There are a number of very good publications out there that will give you the power consumption of the latest device, the features in the newest software release, the data rate of the fastest SerDes transceivers, and the number of marketing gates in the next upcoming FPGA family with no analogies, stories born of the author’s childhood trauma, or farcical analytical diatribes. We even read some of those publications ourselves. We’re just not one of them.
The second category on writing style involves the complexity and subtlety of the language and vocabulary that we use. This one is tougher. We understand that a large portion of you are not native English speakers. We’d love to be able to communicate our points in clear, clean, elegant, and simple Hemingway-esque prose. However, sometimes simplifying the language takes away important aspects of what we’re trying to say. We’ll work on that. Meanwhile, grab a good dictionary and decode our secret messages. We promise it’s easier than learning VHDL.
Finally, many of you asked for more “how-to” articles and webcasts. We’ll work hard to bring them. The truth is, we want to learn how to do new and cool design tricks, too, so we’ll go on a learning spree and share our best discoveries with you. If you have a good trick or tip, send it in. We’ll probably publish it.
Finally, if you didn’t respond to the survey yet, we’d still like to hear from you. You can click RIGHT HERE and give us your data and opinions right now. It’ll just take a couple of minutes.
Thanks so much for making FPGA Journal a success. Without you, our wonderful audience, we could not exist.