Details are scant and we haven’t seen the filing, but multiple sources have told us that Arista is set to publicly unveil its S-1 on Monday. We expect the filing to be available on the SEC's Edgar database at 6:00 a.m. PT.
We’ll update this story and provide continuing coverage as more details become available.
(UPDATE, March 31, 8:00 a.m.: The S1 was released this morning. Arista turns out to be profitable: For calendar 2013, the company reported net income of $42.5 million, or 72 cents per diluted share, on revenues of $361.2 million.)
In the meantime – here’s our initial take.
Is Arista Networks the first SDN IPO?Arista was founded as Arastra Networks back in 2004 by Andy Bechtolshiem, David Cheriton, and Ken Duda, long before the concept of SDN was invented. So, it’s difficult for me to call a 10-year-old company that started four or five years before SDN an “SDN IPO.” Plus, calling Arista an SDN IPO, at least to me, feels like cheapening the original vision and execution of the founding team.
Arista has, however, embraced and extended SDN with its focus on two areas: a) the use of merchant silicon to deliver previously unheard-of performance and density for large data centers, and b) leveraging and adding to open-source software, like Linux, to create a programmable network operating system. Arista’s leadership in both of these areas foreshadowed today’s interest in merchant-silicon-based white-box switching and SDN’s push to automate network operations, making it an leader in the SDN movement.
So, what do you call Arista’s IPO? It’s the first IPO of a new class of infrastructure companies that are leveraging software to solve a new class of customer needs within the data center — and specifically in Arista’s case, solving the problem of scale, density, and automation for web-scale data centers.
What Does This Mean for Customers?In the short term, the IPO doesn’t change much; however, long-term, a public stock gives Arista the currency to acquire additional software capabilities or to accelerate the hiring of more software engineers in the battle against the likes of Cisco and Juniper. Raise your hand if you’d like to see stronger Layer 3 Routing, MPLS, or firewall capabilities (or apps) from Arista.
What Does This Mean for Competitors?Your well-funded competitor just got stronger; you’ll need to make sure that your products are differentiated and offer unique value.
Expect Arista to accelerate investments in the following areas:
- Network packet brokering: Arista’s current LANZ/DANZ offerings are a bit light on features, though we hear NPB is becoming a significant revenue driver for Arista (think multiple multimillion-dollar deals per quarter). Expect to see continued investment going after the Gigamon, VSS, and Big Switch network-tapping use cases. (Separately, if you are interested in learning how to use SDN for network tapping, attend our webinar with the Open Network Foundation to learn how to use a sample OpenFlow application on Opendaylight (ODL) for basic network tapping. Register here.)
- Policy and Orchestration: Arista needs a story beyond partnering with VMware’s NSX to counter Cisco’s ACI while fighting off upstarts like Cumulus that are claiming better automation. We see through our Wiretap Ventures and SDxCentral engagements how policy and orchestration are becoming key software differentiators. Expect Arista to develop a broader capability here.
- Application Ecosystem: As higher performance processors are included in Arista’s switches, creating a developer ecosystem that leverages EOS is key to building long term differentiation. Maybe, one day we’ll even have the Arista App Store.
Not much. As we’ve written before, Arista was not financed by venture capital and does not serve as model for the traditional venture-based companies in the networking space. See my other posts: one on suitability of networking for venture capital, and one titled, “Should I join an SDN startup?”
What Does this Mean for Employees?We’ve heard that Arista is a tough place to work and many of the long-term employees will become instantly rich. Arista is going to have to a challenge managing the tired and now wealthy employees who maybe interested in “resting and vesting” while motivating a new generation of employees to work just as hard without the promises of life-changing wealth. Though given Bechtolsheim and CEO Jayshree Ullal’s experience managing through such inequalities at Cisco (those who went to spin-ins Andiamo and Nuova and those who didn’t) — if anyone can make this work, it’s them.
What Does This Mean for Investors?I’m on the fence as to how Arista will be as a public company stock, not because of Arista’s execution capabilities, but because of possible unrealistic investor expectations and the performance of recent networking IPOs.
Personally, I wish I had an opportunity to own some Arista friends-and-family stock before the IPO. Although that won’t be happening, as, the odds of any of us landing Arista-friends-and-family shares is likely less than winning Warren Buffett’s billion-dollar bracket.