By Lior Mishan
As you might expect, what grabbed the headlines at last week’s Mobile World Congress were the flashy, high-profile displays. Foldable phones, autonomous vehicles, piano-playing robots, and virtual and augmented reality demonstrations were just a few of the glitzy attention-getters. And of course 5G, which either dominated or lurked in the background of almost every conversation.
But underneath the surface, there is solid effort going into telco and cloud networks to provide the infrastructure to make all these things work with the speed and low latency that everyone is anticipating.
What we saw – in more than 60 meetings with leading vendors, technology partners, customers, and industry analysts – is clear, positive momentum for acceleration. Telecom and cloud providers are clearly in the market for more and better resources for their networks, especially at the edge. Speaking of the edge, it is becoming obvious that edge computing is really a broad term, and one with many interpretations. At the same time, the edge is broadening far beyond the exclusive domain of telcos or cloud providers. It includes players you might not immediately think of, such as tower companies that are deploying their own compute resources and that are allowing developers to store applications in their facilities that can service many telecom operators simultaneously.
Our discussions with a diverse group of industry players reinforced our belief that operators must focus on offering more and better services for businesses. That requires cooperation with vendors who understand how to provide these services to this customer segment. Without that emphasis, telecom operators may soon find themselves in direct competition with vendors such as Cisco and Juniper, who have their own established networks of partners and proxies to sell to businesses. The borders between vendors and service providers are blurring, and providers are coming to realize that they need to strongly differentiate themselves if they don’t want to end up in the background while others steal the spotlight.
Through our discussions, particularly with service providers and partners, we see an emerging consensus that FPGAs are the solution for the necessary acceleration at the edge. We really did not encounter anyone who favored multicore over FPGAs. That continues the trend that we saw developing in the second half of 2018, in which industry preferences are shifting solidly in the direction of FPGAs. In fact, an executive from one of Asia’s largest operators told us, “We understand that FPGA acceleration is the right direction to take in order to power the edge deployments that we are planning over the next couple of years.”
There is good reason for embracing FPGAs. As we point out in our white paper, Enabling the Virtualized Edge with FPGA SmartNIC Data Acceleration, FPGA-based SmartNICs use hardware to effectively accelerate the virtualized networking functions at the network edge and produce efficiencies at the edge. This is just one reason why they are on the way to becoming the de facto standard for SmartNICs. Additionally, these scalable, programmable systems can slot into existing servers and deliver both space and power efficiency, while at the same time enhancing security by enabling network isolation and user segregation to prevent attacks on edge sites and user devices.
We believe that the trends that we saw gathering additional momentum at Mobile World Congress will continue to snowball throughout the rest of 2019, with FPGAs achieving that status of de facto standard much sooner than even we might have predicted a year ago. So while the high-profile headlines from MWC were very impressive, more foundational stories for telecom and cloud providers could be found beyond the glitz, perfectly positioned with Ethernity at the network edge.