THE PANDEMIC-DRIVEN DATA TSUNAMI
On March 14, 2020 my wife and I boarded a plane from Sydney to San Francisco. Earlier that week I participated at the Adapt conference on the Gold Coast. Little did I know that it would be the last in-person event I attended for over a year. Near the tail end of the conference, we heard that Tom Hanks and his wife Rita Wilson, who were staying a few miles away from us, had contracted Covid-19. The world was abuzz with the news and it started to hit home that things were going to be very different going forward. We found out during our flight home that it was the last one to leave Australia. The pandemic lockdown started while we were in the air. All other flights in and out of Australia had been canceled. We landed at SFO 15 hours later and saw that the entire world had changed. We were extremely lucky to catch that flight back home.
When the lockdown orders started in the US just a few weeks later, one thing became crystal clear; digital infrastructure was about to take center stage like never before. Seemingly overnight businesses, local and federal government offices, school districts, colleges, universities, and individuals all over the world were forced to complete tasks online. This started one of the largest growth spurts in our industry’s history. In 2014, IDC predicted a 10x growth in data with 44ZB of data generated in 2020. In March 2021, IDC verified that the world generated 64.2ZB of data, a 16x growth rate. 2020 was an unprecedented year. However, contrary to popular belief, it was not an anomaly. It has become the new baseline. IDC’s 2018 prediction was that the world will generate 175ZB of data per year by 2025. If we apply the same growth prediction to the new baseline of 64.2ZB in 2020, then the world will generate over 255ZB of data by 2025. That is assuming that we continue on the same trajectory with the same number of people and machines doing the same things. That is not the case. People will not be streaming back into offices full time. It will be a hybrid of remote working, less travel and more highly interactive, broadband heavy, compute-intensive online activity across the board. Keep in mind that the numbers above assume approximately half the global population will have access to the Internet via broadband. This also predates new investments such as the US’ plan to provide 100 percent broadband coverage for the country. Cybersecurity Ventures predicts that two billion more people will be going online by 2022 in China, India, Africa and LATAM alone. Couple that with the acceleration of smart cities, edge deployments and rural area broadband connectivity programs and we could easily exceed 255ZB of data by 2025. Contrary to popular belief, this data growth is not initiated by humans directly. The Internet 4.0 is being built for machines to talk to each other to enable humans to be more engaged and productive. The number of things connected to the Internet in 2021 will be 46 billion and is predicted to exceed more than 125 billion by 2030. The number of sensors behind those things will exceed one trillion. The advancement in distributed edge computing and quantum computing in the core will enable massive processing of data sets at a scale we have not seen before. The barriers are being broken down and there will be incredible leaps in capabilities in the first half of this decade. The adoption and use of technology will continue at a staggering rate, driving this unprecedented demand.
This is the new normal and Digital Infrastructure will need to handle this data tsunami. Tony Bishop and Dave McCrory from Digital Realty eloquently outlined this phenomenon in the last edition of InterGlobix Magazine’s cover story on Data Gravity. They state that the continuous data creation lifecycle underpins Data Gravity based on this equation: (Data Mass (DM) x Data Activity (DA) x Bandwidth) / Latency2 = Data Gravity. More data, more use, more bandwidth with faster access closer to you draws more data to you like gravity attracts planets or the stars. The data tsunami has been building for years and the pandemic just made it bigger.
SUPPLY CHAIN STRAIN
At the first Infrastructure Masons virtual event on Earth Day 2020, I had the pleasure of having a conversation with Noelle Walsh, the Corporate VP of Cloud and Innovation at Microsoft. That event was the launch of the iMasons’ Sustainability Vision—Every Click Improves the Future. In that interview, Noelle shared a stat that blew everyone away. Microsoft had turned up 100MW of new capacity across their portfolio in two weeks to support the surge of demand in their online services. That was the largest turn up in the shortest timeframe I had ever heard of. It highlighted a number of things for many of us. First, the demand was real and will not slow down. Second, the supply chain was going to be constrained unlike any other time in our history. Third, digital infrastructure personnel were critical and had to drive much of this remotely.
Microsoft is one of hundreds of thousands of companies competing for the same resources. All of them are counting on partners who have to operate at diminished capacity because of the pandemic restrictions on the number of people allowed in manufacturing and assembly plants, construction sites, and on-site turn-up services. What happened in 2020 made all of us in the digital infrastructure industry very proud. We found a way. Teams all over the world delivered the capacity that enabled the massive underlying growth that every person on the planet counted on.
Watching the professionals in our industry rise to the occasion was awesome. We had numerous iMasons’ virtual events throughout the year showcasing people, companies and their accomplishments. The world counted on our community and we delivered. With so many people impacted by the pandemic, we also reflected on how lucky we were to be in this industry. We all had difficult challenges in planning, logistics and delivery, but we were employed and involved in one of the most important deliverables in history. As I have said for years, Digital Infrastructure is the lifeblood of society. Without it, things slow down and can even stop.
Another important element that emerged during the pandemic was the need for enhanced technology. Teams had to rethink how to address global-scale challenges, while providing new features for an online population demand that grew orders of magnitude almost overnight. Zoom went from 10 million daily meeting participants in December of 2019 to almost 300 million in less than a year. Amazon logistics shipped 1.9 billion US packages in 2019 increasing to over 5.1 billion in 2020. Google grew from 1.2 trillion searches in 2019 to 1.9 trillion in 2020. Facebook, the largest social network in the world, went from 2.6 billion active monthly users in 2019 to 2.8 billion in 2020. There are hundreds more examples of this unprecedented growth caused by the pandemic. The key message here is that this demand required significantly more global capacity and radical innovation to serve it.
This led to many discussion at the iMasons’ Board, Advisory Council, and Local Chapter meetings. Those discussions led to an expansion of iMasons’ strategic initiatives. Since our inception in 2016, we have been focused on enhancing Education opportunities, championing Diversity & Inclusion (D&I) and inspiring Sustainability through member engagement and programs that raised awareness of our industry. Our community has been making great strides in all these areas but there was an element missing. Our members wanted a platform to collaborate on technology innovation, but we knew we had to do it in a way that would protect the inventors and innovators.
Our members leave their companies at the door and connect as individuals, while still adhering to their company policies. As we started to brainstorm on the strategy and structure of the new Technology Committee, we realized this was a very different beast than our other initiatives. Education, D&I and Sustainability are non-competitive social issues we all face. There is limited IP risk. Participants are able to openly share their ideas and collaborate on solutions that can be equally utilized by any member. Technology has different challenges. Some ideas can be openly shared and collaborated on, while others are competitive differentiators and inventions in development. The challenge was how to innovate together, while protecting the ideas of inventors and innovators in our industry. Remember how disruptive it was for Facebook to share their open data center and open server designs? This was always considered a closely held secret and made each company’s digital infrastructure unique. The outcome of the open projects revealed that many company’s closely held secrets were forced to measure up against the open standards. Company IP is a competitive advantage in a highly competitive market. Yet, at iMasons we believe we need to help accelerate innovation to serve the increasing demand I outlined at the beginning of this article. This brought me back to a number of successful innovations I’ve had in my career through collaborations with talented and passionate people, while still protecting IP.
COLLABORATION & INTELLECTUAL PROPERTY
My first job in Digital Infrastructure was in 1989 as a component-level debug technician on the manufacturing floor of Sun Microsystems. I debugged failures on Sun server hardware. During my troubleshooting, I found two methods that could narrow down failed components faster. That led to the creation of two hardware solutions and patent applications for them. This was a great learning experience and helped me understand what IP was and why it was so important for companies to protect it.
Fast forward to 2007 and I had an even cooler experience. Mike Ryan and I were able to collaborate with Andy Bechtolsheim, one of the founders of Sun, on a project to help cool supercomputing facilities. Project Greenhouse was created and led to the filing of a patent on the concept (US20080291626A1). We took that further in collaboration with Liebert to incorporate their XD technology in the solution. Jack Pouchet, who is now with Natron Energy, brought a team of Liebert leaders to work with my team to drive super-efficient density for high-performance computing that utilizes pressurized cold aisles. I still fondly remember those working sessions with such great minds. We were able to work together on solutions that boosted both companies. Many of the results from this were shared in solution briefs for customers at our data center openings.
Shortly after that, I met a few people who have been hugely influential in my career: Eddie Schutter and Olivier Sanche.
Eddie Schutter, a founding iMasons member and board director, was with AT&T at the time. AT&T had IP related to data center modularity and some forms of curtain-based recirculation management. My team at Sun had been driving modular data center projects for highly efficient, dense compute deployments through our POD architecture and collaborating with engineering on new data center product offerings. AT&T and Sun met in San Jose to discuss collaborating on efficient modular designs that could be deployed in multiple containment models. Within a few weeks of meeting, Sun’s CIO Bob Worrall, gave AT&T’s CIO Thaddeus Arroyo a Sun Black Box, which was one of the first modular data centers designed into a 20-foot shipping container. The result was disruptive for AT&T as they had to modify their existing modular designs to compete with the Sun black box designs. Eddie and I spent the next year collaborating on implementing the benefits of the Sun POD architecture into the AT&T facilities, which adjusted for NEBS (Network Equipment Building Systems) and other requirements. AT&T’s customized version was called the HD-GEM, which optimized their equipment deployment, shrinking it to half the space at less cost per kilowatt. This was another great example of companies discovering and innovating together.
My team at Sun was having a blast innovating on infrastructure designs as we consolidated data centers globally, creating some of the densest and most efficient deployments in the world at the time. Following what I had learned in the past, we started working with the Sun legal team to look at what patents we could file. At that time they informed us that they didn’t believe what we were building could be patented as new data center designs were difficult to differentiate from existing patents. With that knowledge, we decided to publish two blueprints to share our learnings with customers and the industry: The Role of Modularity in Data Center Design and Electrical Design.
I met Olivier Sanche in February 2008. Terri Jordan, who was the VP of infrastructure at eBay at the time, asked if I knew Olivier as she was considering him for a role at eBay. I said we had connected on LinkedIn, but I had not met him personally. I reached out and discovered that Olivier’s boss had suggested that he connect with me based on the work I had done with AT&T. I learned that Eddie and Olivier had worked closely together on the SBC/AT&T merger. A week later we met in London. Olivier landed the job with Terri and we became fast friends. He moved to the US a few months later to drive eBay’s data center portfolio. Over the next two years, our teams got together in design charrettes that included debates where Olivier and I took opposing sides in front of our teams. Olivier asked to license or use many of the ideas we developed in these collaboration sessions with Sun, AT&T, Switch, and others for his own data center build-outs.
In 2009 Oracle successfully bid to acquire Sun Microsystems. I decided to take the next step in my career and pursue new opportunities. At the same time, I found out that Olivier was being courted by Apple. On a phone call while I was on vacation, we realized that there could be a unique solution here. Over the next two weeks, Olivier accepted the Apple job and I took his position at eBay. We overlapped by a week where he handed off the portfolio to me. I immediately dug in to apply the principles of standardization and scale.
Over the next seven years at eBay, we built data centers that outperformed the deployments at Sun. We achieved PUEs all the way down to 1.018 in Phoenix through modular data centers developed in close collaboration with Dell and HP. This had great results in our owned data centers, but it was a challenge for colocation. Our racks were too dense. This led to a collaboration with Rob Roy, CEO and Founder of Switch. They had no problem powering and cooling our racks that were reaching 50kW each. I remember my first tour with Switch in 2009. I thought we built cool data centers, but Switch took it to an entirely new level. For those of you who have ever toured their site, you’ll know what I mean. It’s like a Mission: Impossible movie set. The black iron forrest, bright red TSCIF hot aisle containment, and tri-power electrical design are tightly integrated, but still allowed flexibility for any rack density anywhere. I remember when Olivier called me during his first tour of Switch in early 2009. He was blown away just like I was. Rob is meticulous on how the data center looks but even more so on how it functions. You can’t achieve 100 percent uptime for 15 years without solid design and operations. In many ways, Rob was ahead of the industry in efficiency and designed redundancy, and successfully patented his ideas, while helping customers to experience the benefits. Trust me when I say that we pushed those designs with the eBay deployments. Our monster racks were some of the densest in the industry, but they hummed along at Switch with no problem. The collaboration between all of our companies during this time was incredible. It yielded great fruit for all involved. Sadly, my friendship with Olivier was way too short as he passed away from a heart attack on Thanksgiving in 2010. What we were able to do together is still an inspiration to me to this day.
These experiences reminded me of the many inventors and innovators in our industry. The largest patent holders are companies like IBM, Samsung, Canon, Microsoft, and Intel with hundreds of thousands of patents among them. At an individual level, there are inventors like Sun Microsystems’ Founder Andy Bechtolsheim, with over 35 active investments and 11 exits to date, including being one of the first investors in Google. Andy has enabled over 10,000 patents through Sun and Arista alone. Switch’s Rob Roy has over 500 issued and pending patents to his name, followed by Marian Croak from Google (formerly of AT&T) with over 200 patents to her name. James Thomason, currently the CTO at EDJX has been part of 13 startups in his career including one acquired by Dell. He, too, enabled significant patent applications across the various startups he was involved with. He became the second highest patent submitter at Dell after the acquisition.
This brings me back to my main point. For us to accelerate the pace of technological development, we need to enable discovery and foster development. The challenge is that in this day and age, it’s almost impossible to not be influenced by something you see, hear or feel. Think of music. Rearranging a few chord progressions can make a new song, but the original creation of that genre is an entirely different thing. That is the key difference between invention and innovation. Invention is the creation of something completely new, like the car replacing the horse and buggy. Innovation is introducing a concept of use of an idea or a method. Karl Benz was the inventor of the modern car in 1886, but the Ford Motor Company innovated on that idea releasing the Model T in 1909 taking that invention to scale.
Today, I’m the CEO of Virtual Power Systems (VPS), a startup company that is disrupting the way in which we manage power in data centers. The intellectual property that’s been created to enable this brand new wave of efficiency started back in 2014 when founder Shankar Ramamurthy applied a software-defined methodology, like compute, storage and network, to the power chain. Software-defined was someone else’s invention. The application of this methodology to the power chain is the VPS innovation.
IMASONS TECHNOLOGY COMMITTEE
My experiences over the last three decades have given me an appreciation for the importance of balancing collaboration and protecting intellectual property. iMasons’ leadership has had deep discussions on how to achieve that balance in our new tech committee.
The good news is that we have some seasoned executives stepping in to help. Eddie Schutter has accepted the role of executive sponsor for the technology initiative and Christian Belady, GM of data center innovation at Microsoft, has accepted the role as vice-chair of the technology committee. We have aligned on a vision that will guide our journey.
The iMasons technology vision is to ensure that every innovation is realized. The technology committee will harness iMasons’ collective experience and resources to discover, guide and accelerate digital, infrastructure while keeping an eye out to protect the inventors and innovators in our membership and partner companies.
If you would like to be part of this journey of discovery, invention and innovation, become an iMasons member at https://imasons.org/join. If you’re already a member, join the technology group through our member platform and dive in.