On the Nature of Innovation
I recently had the opportunity to chat with Soni Jiandani about the nature of innovation at Cisco. For those of you who know her, you know she knows a thing or two about bringing innovation to market. Soni is currently a VP in the Server Access and Virtualization Business Unit. In a prior role, she was Vice President and General Manager of Cisco’s LAN and SAN switching business unit within the Data Center, Switching and Wireless Technology Group, where she was responsible for the industry leading Catalyst modular switches and a comprehensive portfolio of intelligent SAN switches.
Omar Sultan: Soni, you often refer to the “Cisco Innovation Edge” – What does this refer to?
Soni Jiandani: Specifically this refers to our ability to introduce technology innovations more quickly than the rest of the industry that deliver quantifiable benefits to our customers. This really speaks to our culture of being a customer-centric company. Throughout our history we have maintained a close collaboration with our customers in order to understand their current requirements and vision for how technology can benefit their business in the 3-5 year time horizon. By listening and then applying our technology expertise and experience of over 2 decades, we have been successful in driving many innovations into our architecture through custom silicon development. At the same time, we are often leading the charge to drive these innovations through the standardization process and providing engineering support for plug-fests and industry-wide interoperability efforts. We do this because we believe it is in the best interests of the customers, the industry and Cisco.
A recent example of this is Fibre Channel over Ethernet, a technology which is helping customers to save up to 50% on their data center infrastructure and operation by consolidating equipment and networks. We were the first company to introduce FCoE switching products in 2008 in our Nexus data center switches. At the same time we were helping to lead several industry-wide efforts which resulted in Lossless Ethernet and FCoE becoming official IEEE / INCITS standards in 2009. Our custom silicon development allows us to introduce innovations more quickly to our customers, allowing them to benefit through cost savings and more agile infrastructures. Once these innovations become standard and required by customers, they may then make it into merchant silicon; but often this can be several years later. That’s a significant “Cisco innovation edge” that delivers real value to our customers maximizing investment protection. We are on an innovation treadmill here. As we standardize Cisco innovations, we are constantly investing in the next set of innovations.
OS: There are a number of competent companies out there that make merchant networking silicon, so why does Cisco continue to commit R&D resources to make our own chips?
SJ: It goes back to that “Cisco innovation edge”. We want our customers to benefit first from new technology innovations. Since we develop our own ASICs, we can directly execute on what our customers are telling us. Without this ability, we would be limited by what was available on the market. That hurts innovation and ultimately hurts our customers. That’s why we maintain a large R&D budget. We significantly outspend our peers in the switching space in order to maintain that innovation edge. In fact, we estimate that we have delivered more silicon in the data center switching space in the last 3 years than all of our competitors combined. Based on current investment profiles, we expect to maintain this leadership position for the foreseeable future. It really goes back to providing better solutions for our customers. Let’s take switch port density as one example of helping customers save money. Today, we are delivering 96 Unified Ports (any port can be configured as 1GbE, 10GbE, FC or FCoE) in our Nexus 5596 UP when competitors basing their products on merchant silicon can only talk about delivering 64 ports – not unified ports – sometime in the future. It comes down to driving innovation rather than following, often at a much slower pace.
Another key example is that our customers, both enterprise and service provider, need networks that scale. So they require intelligent network services such as multicast, security, QoS, predictable low latency and low jitter, lossless 10GbE over which to consolidate LAN, NAS, iSCSI and SAN traffic, as well as virtualization within and across data centers. Merchant silicon-based offerings may lack these features or be limited in the depth to which they can deliver them over a consolidated DC wide fabric. That means customers trying to build networks with merchant silicon based offerings may find themselves limited in their network’s scale and scope.
OS: Soni, in your view, what is the role of merchant silicon in our product portfolio–when do we use it and why?
SJ: One of the stated elements of our corporate culture is “No Technology Religion”. The underlying concept is that we have the freedom to choose the solution that allows us to best meet our customer’s needs and not get locked into ideological silos. Cisco continues to invest and drive innovations and standardization efforts with the development of our own ASICs because we feel it allows us to deliver better solutions to our customers. However, we do take advantage of merchant silicon in specific use cases, when our customers ask us to solve select requirements. Looking ahead, this approach will not change.
OS: Can you give me example of when Cisco might use merchant silicon?
SJ: Cisco does use merchant silicon for various components in our products. However, basing an entire switch portfolio on merchant silicon is rare. It only happens if a market has a very specific, straight-forward requirement, with none of the usual complex technology trade-offs that can best be solved with custom silicon and broad portfolio of offerings.
OS: So, what would differentiate a Cisco switch based upon the same merchant silicon that other companies may be using?
SJ: When customers buy a switch, they are investing in an integrated system including software, not just a chip. They are also investing in the company that developed that integrated system. So, they evaluate the full solution and investment protection that will be delivered by that company. In our case, we have a number of areas that create meaningful value for our customers. One example is our common data center switching operating system – NX-OS. Already deployed by over 10,000 customers, it draws on over two decades of networking expertise and a decade of storage expertise. NX-OS has proven itself as a platform that delivers the “Cisco competitive edge” with industry-first features like Virtual Port Channel (VPC), and FabricPath both which enable our customers to scale DC wide fabrics with all the attributes of network services including predictable low latency, security, QoS. This functionality is built on our Nexus platforms that support industry standard TRILL within our ASICs. And we continue to innovate and evolve NX-OS, while providing investment protection for our customers all along the journey.