What I learned from Mobile World Congress the last three years

MWC17 just wrapped up, and with this one, I started to recount the path of my last three MWC and how Ericsson as a company and ICT as an industry has evolved over the last three years. Ericsson went full steam ahead on cloud in the summer of 2014 with the new Cloud Business Unit, so the first MWC we showed up as a new business unit was MWC15. The most common questions I got back then was why does Ericsson want to do cloud ? Where is your credibility ? Stick to the Gs. What happened in 2014 was when the new BU started, we began the work with Intel on Rack Scale Design, a new, efficient, composable and disaggregated way to deploy infrastructure. Why do this ? If one should learn about Ericsson's heritage, we have been deploying global infrastructure for the last 140 years. Our first generation of connecting locations meant our infrastructure was wood and copper cables and we had to do it in every corner of the world. Our next phase of mobility and connecting people started in the 1970s and that meant the infrastructure was Radio Base Stations. With Communications and IT merging we now call that infrastructure Cloud. This is why Ericsson is committed to this space, for us, doing a globally distributed infrastructure is business as usual, standardizing is embedded in the Swedish DNA, one needs to look no further than H&M, Ikea, Ericsson, Scania etc. and realize Sweden, for a population that just recently celebrated having gone past the 10 million population mark is remarkably great at standardization and industrialization efforts around a product then do a global rollout of said product/service. So the big message we drove in MWC15 was Ericsson is coming in with the new infrastructure named Hyperscale Datacenter System and the response I saw was confusion with a dabble of excitement.

MWC16 was a different, after working on the product for another year, we managed to showcase our demo products. So the questions I received in MWC16 was quite different than the year before. The questions went from why to how, instead of asking me why we are doing this, it evolved into how are you doing this ? The questions were, how are you doing this ? How will you compete with the slews of IT vendors who all says they too have the latest and greatest and we have been doing business on the IT side far longer with them, how do I know you can deliver ? Our answer from the business side is simple, it is precisely because we are not an incumbent in the typical IT space that we are able to invest in this next wave of infrastructure, we are a 140 year old disruptor, just as how OCP is able to disrupt, how AWS is able to disrupt, how Apple is able to disrupt, we are coming into the Data Center Infrastructure space with a fresh pair of lens and our own history of tears in the operations of it. From the product side, we were very clear that the challenge today is no one is looking at the TCO part of that infrastructure, we all have silos operating ops on their own and we have the nice power point presentations that makes architecture look clean and cables disappear, what is badly needed is an infrastructure that can adjust to your business, mold to your operations and be agile enough to deploy while providing the security and peace of mind for your business continuity. We made our focus on three distinct aspects of HDS, first, hardware disaggregation is a key to breaking down system level products to further components, this meant working with Intel at the silicon level to jointly discover what the future of disaggregation looks like and how it all composes back together. Second, optical, now that we see data center network speeds starting to go in the 100G realm, how does one manage the pool of disaggregated resources ? To connect at the rack level ? At the board level ? Together with our brothers from the Radio side, we made sure optical was a focus for us and the end result of that would be a bigger pool of hardware that one can manage inside your data center. Third, all of this breaking up through disaggregation and connecting via optics sounds good, but what about my application ? How do I consume the right size of CPU, memory and disk so I can guarantee utilization is up, apps are deployed in a performant level and services have fast time to market ? The answer is a Software Defined Infrastructure. Our software focus has always been to manage infrastructure at the component level, collect data, compose and provide intelligence at the deployment stage and becoming that thin layer between bare metal and hyerpervisor/OS world. The conversations around those particular hows defintely provided our guests at MWC16 some new food for thought and we discussed at length how their respecitve organizations viewed operations, business and technology. MWC16 generated a lot of conversations around proof of concepts, lab works and how we can solve real world business problems together.

Now comes MWC17, where we finally can showcase the work we have been doing the last 2 and a half years, our team today, has a readily available stack of compute, storage and network products that can run NFV, Enterprise IT and Cloud deployments, all stitched together through our software defined infrastructure. It was with great excitement that I went to Barcelona, because this MWC represented the culmination of so much hardwork from those of us in Sweden, this was our coming out party. No longer are we doing demos and power point presentations, we are talking real life deployments and delivery schedules, we are discussing workload on boarding as well as aligning to a world of distributed infrastructure which serves as the baseline for the coming IoT era. We showcased our full rack solution where people can order today, we also showcased our Open Compute Project(OCP) product that is in the works, given the industrialization nature of Ericsson, our thoughts on OCP mainly revolve around power, optical standards and software. Last but not least, what a fully disaggregated, 5G connected, Software Defined Infrastructure with full robotics service capability looks like in 2020, this is what the distributed cloud infrastructure that manages Internet of Things looks like in our minds. The last three years was monumental in establishing our business as one that is forward looking, innovative in technology and disruptive in business, within 3 years, we built an entirely new product portfolio, expanded our global sales footprint and are now in 26 customers and going live this year. The work is just getting started, great job from the Ericsson family and what a wonderful time to be a 141 year old disruptor.