Interview with Paul Nikolich
Paul Nikolich has been chair of the IEEE 802 Working Group since 2001. In this interview with Ethernet Alliance chair Peter Jones for The Voices of Ethernet oral history archive, Nikolich describes the importance of standards development and how it is an opportunity to take an idea and enable it to be deployed widely on a global scale with Ethernet as a prime example.
“Fifty years, it’s been dramatic…the Ethernet group has been tremendously flexible in developing the interfaces for a wide range of target end applications very, very successfully.”Paul Nikolich
Learn More about Paul Nikolich
- LinkedIn profile
- IEEE 802 LAN/MAN Standards Committee
- IEEE Entrepreneurship
- Paul Nikolich | IEEE TV
- University of New Hampshire’s Broadband Center of Excellence
Peter Jones: Welcome back to Voices of Ethernet. Today we’re talking to Paul Nikolich, one of the busiest people behind Ethernet’s rise through his account and experience with a central group of technologies history, this is the LAN/MAN Standards Committee otherwise known as IEEE 802. Thanks for joining Paul. Can you introduce yourself please?
Paul Nikolich: Sure. Thanks, Peter and thanks for having me on. I really appreciate the opportunity to give you my perspective on what some of the wonderful things we’ve done in 802 specifically in the field of Ethernet. I’m self-employed. I am an advisor to several small tech companies. I’m also an angel investor in tech companies and on the board of some nonprofits and companies. And I’ve been chairing 802 for the past 22 years, it’s hard to believe. But yeah, that’s my focus, my focus is on helping 802 develop standards and companies to develop products and services that’ll be useful in the marketplace.
Peter Jones: So, Paul you’ve been chair of 802 since 2001, right? This is a long time. So, when and how did you become involved?
Paul Nikolich: Yeah, so I first got involved in 802 when I joined a company called Interlan which was developing Ethernet interfaces back in the ’80s. The original Ethernet interfaces that were for thick coax and thin coax and the hot new physical layer back then in the late ’80s was twisted pair wiring, which existed in buildings, you didn’t have to pull new cabling in order to network your computers together. So Interlan hired me. My technical specialty is analog and RF design and they hired me to develop their 10Base-T interfaces. 10Base-T was a new project in 802 specifically designed to carry Ethernet packets over the twisted pair wiring I just mentioned, the telephone wiring that existed in the walls. And that’s how I got started, I got started as a regular participant helping to develop the 10Base-T standard which we completed later that year. And I found it fascinating, the whole process of standards development and building consensus and working together with natural competitors but working together cooperatively was a very, very intriguing opportunity for me. And I really enjoyed it and I gradually assumed more and more responsibility within 802 going from the 10Base-T project to leading a conformance test project to becoming the vice chair of 802.3 then becoming the vice chair of 802 in 1996 and then the chair of 802 in 2001. And I never believed that I would be still chair of 802 in 2023 yet here I am.
Peter Jones: You must have done a good job.
Paul Nikolich: That’s it for me, I have made it clear to everyone that 2024 when my term is up, I’m done, we need to get some fresh young blood into the leadership of the organization.
Peter Jones: So, standards development is often seen as this mysterious thing that happens behind closed doors. So what do you think is the thing that’s most commonly misunderstood about standards development?
Paul Nikolich: I think what’s misunderstood about standards development is that it takes too long. People and executives say, “Well, you know, if you’re going to develop this standard it’s going to take two to three years and why does it take so much time to do that?” And the fact of the matter is that if you have a successful project that’s meeting a need in the marketplace you will have a lot of people engaged in the participation of that project to develop that particular standard for in the case of 802 and 802.3 the Ethernet interfaces. And it takes time to build consensus, you have a lot of individuals bringing different perspectives, different ideas on how to accomplish the project’s goals, and the chair of the project has a tremendous job in terms of getting people with diverse opinions to eventually come to a common understanding of how to proceed and then moving on. So, what’s most misunderstood is the time it takes to build consensus and the value associated with the consensus, once it is arrived at. The value being you have the benefit of dozens, if not hundreds of people’s experience making sure that the technical specification is of the highest possible quality so that when manufacturers go out to build components, silicon, systems, software, it will easily interoperate and flawlessly interoperate despite who’s manufacturing the platform. So that’s I think the least understood aspect. We understand it Peter and I and everybody that’s in 802 certainly understand it, the value of it and the necessity of it, but the outsiders, they don’t get it. It’s like, “Why can’t you move faster.” Democracy is slow.
Peter Jones: Well, that’s also providing a document that everyone can implement and everyone gets right, everyone can get right is hard. So as 802 Chairman, right you oversee a lot of standards happening at once, right, so how do you manage that?
Paul Nikolich: By having an excellent team around me. The way 802 is organized, it’s organized into working groups and within those working groups there are projects and within those projects there’s dozens to hundreds of people working on the individual projects and most of them are very dedicated and very skilled, experienced…
Peter Jones: Very opinionated.
Paul Nikolich: Very opinionated for sure, which is good, you want opinions, you want different perspectives. But that’s how I’m able to oversee all of the work is by having excellent people around me from the upper management, from the executive committee level down all the way through the active participants.
Peter Jones: So this is the classic delegate– delegate correctly is the way to get good results.
Paul Nikolich: Right, right. One of the benefits of being the chairman is you get to appoint several of the executive committee members. So I get to appoint the two vice chairs, the two secretaries, the treasurer, members emeritus and the working groups that I mentioned earlier like the 802.3 working group and 802.11 wireless LAN working group, we’ve got five different, five, six, seven groups, working groups and technical advisory groups, they elect the chairman and the vice chairman are part of the executive committee. But all of those people, every single member of the executive committee is a really experienced knowledgeable individual in the field of standardization and strongly endorse and adhere to the principles of openness and transparency that has made 802 so successful.
Peter Jones: Okay, so what do you think is the most important lesson that you’ve learned over your time?
Paul Nikolich: Surround yourself with good people and the better they are, the more skilled they are, the better it is for the organization and the better is for me. I love the fact that most of the people that I’ve got reporting to me are better engineers than I am and better standards developers than I am. I just kind of try to set a high-level direction and they take off accordingly to their own particular market needs.
Peter Jones: Sort of to go on with that a little bit, certainly one of the things that I’ve learned is how important it is to be able to understand other people. To build consensus you have to basically be able to figure out what other people want and need otherwise if you just show up and say, “This is what we should do, you get nowhere.”
Paul Nikolich: Absolutely, you have to be a good listener, much more important than being a good talker.
Peter Jones: And that’s not a skill that comes notably for many engineers.
Paul Nikolich: Well some people have a tough time with it, I agree, but most people eventually get it, they eventually get, “Listen, I may have a different idea than you do, but I really shouldn’t dismiss your idea just because it’s coming from you, I should understand that it is that you’re proposing and why you’re proposing it and how it might dovetail with my particular approach.”
Peter Jones: Yeah, this is the, “How do we get to a we, not an I?”
Paul Nikolich: Right, right. It’s a challenge and there have been very big battles over differing approaches that I’ve seen through my time in 802 that have resulted in significant delays that may have been avoidable if people didn’t take such firm antagonistic positions. But it’s just part of the standards development process is that sometimes you’re unable to come to consensus and then you need to figure out what to do. And so what we’ve done in 802 is we’ve either terminated a project, if we can’t come to consensus then we’ll just terminate. We have terminated projects. Or we’ll split them off and we’ll get two projects going that are different than one another, sufficiently different than one other such that each camp can take their own tack and eventually the market will decide whether or not one is more important or better than the other.
Peter Jones: So one of the things I’ve tried to learn to think is, “Okay, this is really important to me, but will any of our customers actually care?” Because sometimes you look back and go, “If it meets the needs and our customers won’t care,” then you say, “What’s the quickest way to get market?”
Paul Nikolich: Well, I mean the customers have to care, I mean you shouldn’t start a project if you don’t know that you have customers…
Peter Jones: Let me say different, right, so we’re going to build this, right, but often you see people want a particular thing. Now if that particular thing is very important, like a particular coding scheme, but the customer’s not going to end up caring than maybe that becomes less important.
Paul Nikolich: Oh okay, yeah. Yeah, that’s true. So who is the customer, that’s a good question, the customer the way you’re defining it Peter is the individual or entity that is going to build the equipment and offer the service and if the individual that is using the service, the data communication service really doesn’t care about the fine grain details buried down deep inside of the infrastructure. I mean what we do inside of 802 certainly Ethernet it’s about the data communications infrastructure. Ethernet supplies the infrastructure on top of which I would say well more than 90 percent of data ultimately flows over, all the wireless data’s got to be connected via wireline communications and that wireline communications is 90 plus percent of the time going to be based on the Ethernet specifications. So, to your point Peter, a fine grain detail with respect to a technology approach the end user doesn’t care about, but choosing the right approach can affect whether or not the platform, the equipment or the service can be economically offered to the end user ultimately.
Peter Jones: That I would agree with.
Paul Nikolich: And deliver the reliability that needs to be made in order to have the service, whatever the end application is, have that end application service operate properly. With choosing forward error correction code, boy we’ve had knockdown drag-out fights in that area. But ultimately the ones that are chosen are the ones that are the best for the end user even though they don’t know about it. And lots of times the driver behind the contention for that particular technical approach can be intellectual property ownership. So, if I’ve invented a code I really am going to try to get it embedded inside of the standard because if I can embed my intellectual property in the standard and it does end up making it into chips and systems there’s a significant, potentially significant large financial benefit to me for licensing arrangements. So, part of it is economically driven as well.
Peter Jones: That makes sense. So, what’s the one thing almost no one agrees with you about?
Paul Nikolich: That I would like the individuals that come in to lead 802 be in their early to mid-40s. I’ve been asking, since I decided and announced that I’m not going to be standing for chair again, I’ve been on a pretty strong direction to try to get more young people involved in leadership positions in 802. When I started in 802 I was in my mid-40s and now I’m in my late 60s. And I should have stepped aside sooner, but I really, really love what I do, I really enjoy 802 and leading 802. I call it a very expensive hobby, I don’t make any money out of chairing 802, it’s a donation to the community and I love it, I just enjoy working with you and all your colleagues and changing the world. We have changed the world, Peter in the time that we’ve engaged in 802.
Peter Jones: So what’s your biggest challenges in 802 over the last couple of years and looking forward?
Paul Nikolich: The biggest challenges I would say is bringing technology into 802 that may not be ready for standardization, but conversations should be occurring within 802 about the potential for those technologies. So, a lot of our participants are very focused on, “Hey, I’ve got a project, we’ve got to just focus on this project and get it done before we move on to the next project or opportunity.” And they’re not welcoming to some of the new ideas that are out there like specifically with respect to machine learning and artificial intelligence and how that might be leveraged in building better networking field, networking equipment and incorporating that into the standards. That’s changed in the past year, the 802.11 group has kicked off an interest group that’s examining the opportunities for machine learning and artificial intelligence in the 802.11 standards. But the 802.3 group hasn’t started that as far as I’m aware. It may be discussed and I’m not aware of it, but the big challenge is agreeing that we got to stay focused on the task at hand yet let’s be aware of the emergence of new technologies that we could leverage to our benefit four or five years from now.
Peter Jones: So this is the looking over the horizon idea.
Paul Nikolich: It’s looking to the horizon, not so much over the horizon, but to the horizon is the way I would state it. So that’s a challenge for the group now. In the past, I mean there’ve been other challenges, really big challenges with respect to because there’s such a great amount of commercial activity in the equipment and the potential for revenue by the vendors that sell the equipment and services, sometimes people are trying to establish unfair advantages. And one of the things that I pride myself on is that we provide as level a playing field as possible for the participants to come in to 802 and develop the standards so that when we detect some behavior that is potentially unfair we’ve had to go in and take steps to eliminate that and that’s been very, very difficult and challenging. And it’s unavoidable because of the economic benefits that may accrue from having that unfair advantage. I like to pride– I think we in 802 generally pride ourselves on the fact that we make choices based on what’s best for the particular project and not what’s good for an individual or a company.
Peter Jones: So, can you talk a little bit about how Ethernet has impacted the world over the last 50 years since we’re in the 50th anniversary.
Paul Nikolich: Fifty years, it’s been dramatic. It has been a fundamental enabler of communications between endpoints computers that are endpoints. The computers back when it started were big huge hulking file cabinet-sized things as opposed to now there’s dozens of computers in a single chip. And Ethernet has been the essential connection between those computing entities and it’s grown as the technology has developed the Ethernet interfaces have developed to enable all sorts of endpoints to be connected. And depending on the requirements of the endpoints the capacity of Ethernet has been tremendously flexible. I mean when I got started ten megabits per second was a really big deal and now we’re working on a project in 802.3 that’s looking at 1.6 terabit potential interface. And in addition, and this is the flexibility and how 802 is able to adapt to changing market conditions, in addition to providing faster physical layer speeds, as the industrial Ethernet market for instance has emerged as a big market opportunity they don’t need gigabit speeds, what they need is they need to have twisted-pair multi-drop environment high reliability so that when you connect your machines in your factory that is controlling robots you’re able to cost-effectively connect them and we are developing time sensitive networking technology that guarantees that packets will be delivered when they need to be delivered so that you could have real time control over these industrial processes. And this is also important for intra-vehicular communications. So the Ethernet group has been tremendously flexible in developing the interfaces for a wide range of target end applications very, very successfully.
Peter Jones: I agree with all that. Are there any particular areas of Ethernet development or is that the case of you love all your children?
Peter Jones: Actually I was going in terms of what’s happening today.
Paul Nikolich: Oh, what’s happening today? Well I think the big areas that are in the pipeline right now that are projects right now are the multi-drop twisted-pair physical layer for the industrial and automotive applications and the 800 gigabit per second 1.6 terabit per second optical interfaces for the very high speed networks that are needed inside of datacenters and in between regional networking hubs.
Peter Jones: That’s interesting because we’re going from here right to– the scope is enormous, the range of Ethernet at the moment is huge.
Paul Nikolich: Oh yeah, right. So not only does it go very fast or at an appropriate speed for a particular application, we also deliver power. We go literally from DC to light with the 802.3 working group and we’ve been going from DC to light with that working group for 20 years now. That’s pretty impressive and each thread has continued to develop according to its particular market drivers.
Peter Jones: Agreed. So, if you were going to look back and say who were the people who’ve helped you along the way now with mentors or people that you’ve modeled after, right, who springs to mind?
Paul Nikolich: Well I would say that the leadership of 802 back when I was a newbie in 1989 were critical to helping me understand the value of 802 standards activities. So, the leadership of the 802.3 working group, the leadership of the project, the 802.3 10Base-T project was led at that time by Pat Thaler who eventually became the chair of 802.3 and became the chair of 802.12 and then I appointed her to be a vice chair of 802 eventually. And whoever was in leadership roles at the time influenced me and gave me guidance, Jeff Thompson was the chair of 802.3 when I was the vice chair of 802.3. Whoever was in a superior leadership position in 802 was a positive influence to me.
Peter Jones: Okay, got that. So, what’s one thing about your professional journey you didn’t expect. And I’m actually sort of guessing it was the whole standards thing because I could imagine you started as a normal engineer type and then your career took this dramatic turn.
Paul Nikolich: Yeah, that’s true and the fact that standards were a critical component in my professional journey, the fact that I was a part of 802 and an expert in 802.3 physical layer and I was also involved in the early phases of the 802.11 wireless LAN before there was a wireless LAN working group, I was a strong promoter of radio frequency data networking. And the fact that standards played such a prominent role in my career and also in the ultimate success of the company that I cofounded back in 1998 Broadband Access Systems, we developed carrier class cable modem infrastructure equipment that was based on the DOCSIS standard and we actually started– DOCSIS is the Data Over Cable Service Interface Specification whose progenitor was 802.14 Cable TV Working Group. We started working on cable modem standards inside of 802 before it moved out into the cable television industry cable labs and that standard enabled companies such as mine, the one that we started to be successful. So that’s one surprising thing. The other thing that I did…
Peter Jones: Hold on before you go to that particular one, I’m reminded that we actually have a contact in common being Rouzbeh going back to the time of the start of that technology.
Paul Nikolich: Absolutely. Speaking about mentors I mean Rouzbeh and I have been colleagues and friends since 1987. I started working at a cable modem company before there were standards back in the late ’80s and Rouzbeh was there as the VP of Engineering. I at the time was convinced that cable modems were going nowhere, so I left and I joined Interlan and I…
Peter Jones: If I recall correctly that was the company Applitek because I was involved in the same one.
Paul Nikolich: Yes, yes, it was Applitek and yeah I left Applitek, Rouzbeh stayed and he and I remained friends and colleagues and he stayed in the cable modem side and founded a company called LANcity which was this subsequent company to Applitek and he helped create the market for cable modems with the cable operators and I helped him out a little bit in that company and also he was very active in the data over cable service interface specifications, developing the standards there. So, he and I have been working closely together for geez, a long time, close to 40 years and we’re really good friends. I’m a partner in ventures where Rouzbeh and I together will decide whether or not we would like to invest in some early stage companies.
Peter Jones: So any final thoughts, Paul?
Paul Nikolich: No, I appreciate the opportunity to give you some thoughts about my involvement in 802 and the value of 802 and the impact it’s had. It’s something that I encourage as many people as I can in our industry from students through professionals to engage in standards development because it represents an opportunity to take an idea and if it has real value to enable it to be deployed widely on a global scale. So I encourage people to always consider the opportunities that exist for bringing their new ideas to market through standards.
Peter Jones: Right, so Paul thank you so much for your time today because I really enjoyed hearing some of this because every time I do one of these I find out something new. So to our enthusiast watchers, go take a look at the Voice of Ethernet’s section on Ethernet Alliance or on the YouTube Channel, take a look, tell us what you think. Thank you very much.