Where did you hear that?
The ongoing breakdown in the advertising model leaves engineers and executives on their own for information gathering; expect to put in more hours sifting through data.
DACeZine sat down to discuss the future of information gathering with Limor Fix, principal engineer and associate lab director at Intel Research Pittsburgh and general chair, 45th DAC; J.L. Gray, consultant at Verilab who runs the Cool Verification Blog; David Maliniak, EDA editor at Electronic Design; and Gabe Moretti, managing editor of Gabe on EDA. What follows are excerpts of that conversation.
By Ed Sperling
Q: With the traditional business media collapsing around us, where will engineers and executives go to get their information?
Maliniak: When I first started, there wasn’t any other way for engineers to get this kind of information. There was no Internet, no cell phone—none of this infrastructure existed. We’ve been marginalized by the Internet. The question we’re grappling with is how to make us relevant in new ways. The tech industry sees us increasingly as a middleman they would like to bypass.
Moretti: You also have to bring in the business dynamics. When it was paper-based—and even through the transition from paper to electronic—papers were family-owned. Now they’re owned by international conglomerates. That implies a tremendous push toward profit, and the easiest way to generate profit in the short term is to cut costs. Sometimes they’re shrinking coverage to what is acceptable, not what is preferable while, at the same time trying to deliver a good product at a lower cost. They’re looking for the pain point, and then they’ll back off. But at the same time, advertising is shrinking because companies are also experimenting with new channels.
Q: So where do engineers go, once the dust settles?
Fix: First of all, we want something else. We are now reading the news on the Internet. We are interacting with people on Facebook. I want it short, I want it on my Blackberry, I want it real-time, and I want to hear conflicting opinions from my social network. I don’t want to read an article that is two pages long that is last month’s news. The players that will survive will be the ones to give the new type of information.
Maliniak: So for you, it’s about how the information is delivered?
Fix: The content is also different. It’s not just the format.
Gray: I go through the trade Web sites and publications. The material there is almost irrelevant to me. Sometimes there are articles that are interesting if I can find them, but I don’t browse through the sites all the time so I usually can’t find them. I look for other people I know that have blogs.
Q: Do you trust blogs more than professionally written stories?
Gray: If I’m writing a blog about back-end tools, you wouldn’t trust me because I don’t know anything about them. If I’m writing about verification, then people do trust me because they know that I’m writing about something I’m doing at the time. People know that’s my background. Journalists are writing about something they don’t really understand.
Fix: There are a few trusted names. I’m so busy with my work, I can’t tell you how many different articles I read, but there are a few names I know and trust. But when I go to a conference, I hear someone on a panel. When I go out into the corridor, I hear a little bit more feedback. If I hear it once and I hear it three more times, I trust it. That’s the way we think now. I would like to have an article where enough people comment.
Q: So what you’re looking for is triangulation of information?
Q: Is that something the press has done traditionally?
Moretti: The press has always had letters to the editor, which were the only way to communicate or establish a different viewpoint. [CMP’s] DesignLine family of Web sites is trying very hard to use the Forum section. But engineers are really busy, and they’re afraid of communicating, probably because they feel their managers might get upset if they say something that is proprietary, or they might be misunderstood. They’re more absorbing than giving.
Maliniak: We’ve had many attempts to stimulate back-and-forth discussion, and it rarely goes anywhere. Some sites work, some don’t.
Q: Three years from now, where will you go for information and what will you expect to get?
Gray: I would love to have a way to pull down the topics I want and deliver them to my readers.
Q: Who’s going to be writing those articles?
Gray: One place would be the trade press. A second would be bloggers. I’ve been trying to convince people to write about what they know. Eventually you get a collection of people who know things, and you know they know, and that’s what you want. In verification, one forum that has worked is Janick Bergeron’s Verification Guild. You know that if you go there, he’ll answer your question and he 'll be right.
Maliniak: People who run blogs are not always without agenda. Janick is a straight shooter, but he does work for a commercial entity.
Gray: He’ll take a position, but he’ll be cordial about other positions.
Q: But isn’t a broad-based publication comparable to a supermarket, where you no longer have to go cut deals with each fruit and vegetable vendor? Given what you’re talking about, you will have to work much harder to get your information, right?
Fix: The next wave will be to look at something on the Web that brings you relevant information. Instead of me sitting for three hours looking through everything, within 15 minutes this software will give me a quick look on a certain topic. I’m willing to pay a monthly fee for aggregation.
Gray: How can you charge for something that’s so easy to build? I can spend five minutes on Yahoo Pipes and create an aggregator if I know where all the resources are coming from. That doesn’t cost anything.
Q: Where’s the accountability in the future? We’re talking about disaggregated information sources, and three people may agree and be totally wrong. What happens if you follow it and it ruins your business?
Moretti: That’s the difference between information and data. The difference is the filter and the capability of the engine to give you something you find worthwhile and usable, versus data you have to digest and then transform into something useful.
Q: Do you trust the data, and how regular does it have to be?
Moretti: Limor [Fix] is willing to pay for it, which assumes there is a professional organization behind it. J.L. [Gray] is saying you can build it in five minutes and it runs by itself. The difference is that people very often confuse data with information. We need to be aware of that. Every once in a while there is information on YouTube. But I do not consider that site a source of information: to me its purpose is entertainment.
Fix: If some aggregator would put together information, filter it and I could see the sources of the data and the opinion of the aggregator, I am willing to pay for that.
Q: Unless people are being paid to produce content, there is little incentive to do it on a regular basis. Will that change?
Gray: I’ve been doing Cool Verification for almost three years. I’m getting paid indirectly, because it’s related to my job. If I write 10 articles in a week, I don’t get paid extra. But because I’m a consultant, it’s part of my job. At the beginning, I did it twice a week. It’s not like a daily newspaper. But to be honest, the trade press doesn’t do that, either. If blog sources are giving me something on an irregular basis, the trade press is only giving me something useful every couple months.
Q: What are you looking for in the market? Is it news or opinion?
Fix: It’s a good thing to look for opinions, and I look at whose opinion it is. But I want to know what’s happening more globally, too.
Q: What can be done to fix this in the existing media?
Moretti: We are in a transition phase. Everybody is talking about change. The traditional press is changing because our readers are much more interested in very specific and pointed areas as opposed to the broad engineering publication that covers every bit. But I cannot say I get all my information from just one area. Designers interested in a specific application area, for example, read my publications to figure out the tools to solve their problems. We are still experimenting. The publications that will be successful faster are those focused on very specific goals because they can count on a dedicated readership that will be vested in the content, or at least in the goal, and therefore more willing to support it. As far as the business aspect, we need to learn to look for quality and not just quantity. For example in EDA having five people responsible for a $200 million budget read your publication should be valued more than having 500 designers that can generate only one tenth that revenue amount to potential advertisers.
Maliniak: Things have changed a lot and people are looking for more pointed information. Given the resources we can have in this day and age, I like to think we are doing the best we can. We’re experimenting. When I first joined, I didn’t sign on to be an Internet talk show host, but that’s now part of my job. We are all looking for the right formula that will resonate.
Q: Everybody here has mentioned blogs, but the more successful a blog the more sensational it is. People are flaming other people for the sake of boosting traffic. Is this a problem?
Gray: The media does that already.
Q: The business press has shown much more restraint in researched articles, but in blogs that isn’t necessarily the case.
Gray: That’s true, and I’ve been guilty of that myself at times. The topics aren’t the most exciting, so why not add a little color? Everyone is usually very reserved.
Moretti: I have stopped reading a certain EDA blog because controversy is the primary tool used to increase readership and the content of contributors is often targeted toward generating negative debate. On the other hand, a few months ago I wrote a blog about CPF (Common Power Format) and UPF (Unified Power Format) and it created a lot of problems. I didn’t mean to be controversial, although I was judgmental. Yes, people like entertainment, and there are publications for that.
Maliniak: I did a feature on CPF and UPF, too, and I stopped short of impugning certain parties in that debacle. It’s the usual EDA bifurcation for the sake of bifurcation, not competing on algorithms and tools but locking people in just for the sake of doing that. But it’s never been our charter to be judgmental. A blog format would lend itself to that more, but in print I’d rather present the technical merits of the argument and let them speak for themselves.
Q: With the traditional advertising model eroding and an entire training system for journalists in jeopardy, will the articles that come out in the future be as valid as in the past?
Maliniak: I like to think we, as journalists, do provide a certain sensibility. The training does matter. It goes back to data versus information. It’s easy to get data. It’s the same as with verification tools. You can pile up mountains of data or you can get information about the design. It is about quality.
Fix: There is a big danger of quality going down. We need to find a way to keep the quality, but it has to be presented and delivered differently. Entertainment is not a dirty word. If I’m working 24 hours, the time I have to read anything is usually when I go to a movie and there is a commercial, or when I’m on the plane. It’s my spare time, but I’m working. So it has to be lighter. It has to be high quality, but if it has a short video, I love it. There is value in the different type of communication we expect now. It is a little more personal.
Gray: So what’s wrong with a news source presenting an opinion? There’s a mix. A blog may be focused on an opinion, but it also can be based on expertise.
Maliniak: In a traditional publication, there needs to be compartmentalization. That’s why when you read the newspaper, it has an editorial page with opinion. Next to that are letters to the editor, which are opinion, and commentary, which also is opinion. And then there is news. News is news. It’s objective.
Moretti: I write news, blogs and newsletters. All three are different styles and all require different information. The labels that we put on those things are setting the expectation of the intended readership. They also require a slightly different approach to the material.
Q: What effect does compartmentalization have on the engineering community? Is it important to know the entire flow in EDA, or simply the verification piece?
Gray: There are a lot of different areas of expertise in chip design. I enjoy knowing what the other guys are doing. Some of it is interesting. Some of it is so specialized that I will never know about it. There’s a layer in which you need to be an expert, and many in which you don’t.
Fix: At Intel we educate people about other areas. We educate them about the design flow. There is specialized training and people have to go to it. We are educating people about where the flow starts and where it ends. But people do not have the spare time to educate themselves and go to conferences and attend sessions that do not pertain to them.
Q: Are we missing something by not having people informed on a variety of topics?
Moretti: When I was an engineer, I always wished that engineering students had been required to take at least one behavioral science course and one economics course. They need to play well with the team and understand the fiscal impact of issues. A lot of the problems in submicron design are due to the fact that we have digital engineers that don’t know physics. That is as important as whether a verification engineer has any understanding about the difficulties of a place and route engineer. We are now talking about dealing with DFM problems at the system design level, for example. You cannot leave a solution to the last minute.
Maliniak: In journalism, we were encouraged to delve into everything. I dabbled in everything. It needs to be the same in engineering. There is no digital design anymore. It’s an analog discipline.
Q: Are you optimistic about the future of information for engineers?
Gray: If I was going to be a Python programmer, I could find a lot of information about it. In my job as a verification consultant, I’m not sure how many people can do it, but the number is at least an order of magnitude lower than for Python programming. That may mean I’ll never have a good source of information. How can the press spend time focusing on this? I don’t see how it can be cost-effective.
Q: There’s a difference between understanding the technology, though, and understanding how all the pieces interact on a macro level. You need the most current stuff available. How do you find that?
Gray: I don’t know that it will ever be available for me. The people who do know about it aren’t experienced in writing and it may never come out.
Fix: I have the opposite problem. In formal verification, for example, at the beginning, there were two conferences, five professors. I knew everything and everyone. Ten years ago there were six conferences. I was still able to stay on top of it. Today, I have lost control. I do not know every technology and every tool. I don’t have time to aggregate the information. There are 20 conferences, 60 professors, 300 Ph.D.’s. It’s too big for me.
Gray: There’s a lot of research interest in that area. The information I need is all hidden away inside companies.
Q: Will there be enough journalists to disseminate this information?
Maliniak: I’d like to say, ‘Yes.’ Even if the traditional publishing concerns don’t survive, information will find its way out.
Moretti: We have to look at this as a market problem. Nature is self-leveling. Demand and supply, in the long term, will equalize. That does not mean we’re going to be able to avoid crisis times, where the expectations of who we see as our customer do not match with the tools and the capabilities we have to satisfy them.
Fix: I think we need to talk more. There is a disconnect between the generators of information and the consumers. We have changed our lifestyle and there is a need to re-open everything. But we can find the solutions only by the type of communication we are having now.