Driven by digital innovation, many of today’s network performance requirements are actually much greater than simply delivering voice and video applications across the Internet. Today’s most advanced digital environments leverage hyperconnectivity and hyperscale to enable the super-fast collection, processing, and transfer of astronomically massive sets of data, highly agile instantiation, and unprecedented user consumption of online services.

Advanced pharmaceutical research and aerospace development, for example, require unprecedented levels of processing power and data to support things such as modeling and analysis. Similarly, dynamic e-commerce and enormous multi-user gaming environments rely on hyperscale architectures to deliver unmatched customer experience. And financial institutions, for example, need an environment designed to deliver fast, high frequency trading, while massively scalable Internet services require speed and scale to support customers and maintain service levels.

Digital innovation depends on advanced processing hardware

This tsunami of data and the speed of applications has completely overwhelmed the off-the-shelf processors and computing systems that have traditionally filled the devices sitting in the racks of our data centers and wiring closets. That’s why visionary companies including Apple, Google, Microsoft, and Amazon began building custom processors designed to provide the raw power needed to accelerate the frontiers of digital innovation. These purpose-built, application-specific integrated circuits literally power our modern digital world — and will eventually enable smart cities and smart infrastructures to control things such as transportation systems, on-demand power grids, and responsive manufacturing systems.

The need for a purpose-built security processor

However, the one glaring gap in all of this is cybersecurity. Crack open the most expensive firewall and you will find only an array of off-the-shelf hardware that will never be able to keep up with the breakneck pace of digital innovation and the exponentially growing demand for processing power. Part of the challenge is that firewalls require far more computing power to inspect data, detect malicious content, or prevent cyberattacks than any router or switch.

Unfortunately, while a handful of security manufacturers may have some sort of security acceleration hardware in the works, the vast majority of vendors don’t. And, with one notable exception, none of them have anything in production. This is an issue we simply would not tolerate in any other industry. To meet the security demands of today’s hyperscale, hyperconnected, and hyperspeed requirements, we need security devices designed to keep up. Some of the critical use cases for such development include:

  • Big Data modeling computations that require securing massive amounts of data (>1TB) being delivered through 100G interfaces (also known as elephant flows) at network speeds to speed up discoveries
  • Real time inspection of massively scalable internet services that need to encrypt data at ultra-high rates across multiple 100G interfaces to comply with regulatory data privacy requirements
  • Securing dynamic online gaming and e-commerce sites that offload hyperscale traffic bursts to hardware, while supporting millions of computations per second to maintain user experience and customer loyalty
  • Protecting financial transactions that require single-digit microsecond latency, especially in high frequency trading environments to optimize investment decisions
  • The ability to scale security to support new 5G networks that need to support carrier-grade Network Address Translation in real time and enable rich multi-media content on mobile devices

Other security vendors needed

Without purpose-built security processors, organizations that rely on digital innovation are left with the terrible choice of either slowing down to accommodate security limitations, or not securing their environment at all. In an age when cybersecurity threats are rapidly accelerating, leaving critical processes to run in the open, or at best barely protected by things like access control lists, is a fool’s game that no one should be forced to play.

And once these advanced networking functions begin to be merged into the next generation of smart infrastructures, the lack of viable security tools capable of protecting and inspecting hyperfast, hyperconnected, and hyperscalable environments will result in a crisis we will simply be unable to address.