Cisco has four strategic imperatives for its Nexus 9000 data center product and its Application Centric Infrastructure (ACI) focus: attack VMware/Nicira's software-intensive network virtualization technique; combat the "white box" merchant silicon-based commodity switching trend buttressed by software-defined networking (SDN) and popular among webscale companies like Google and Facebook; enter or stimulate growth in markets where it is not already present or strong - like storage, security and Layer 4-7 services - by virtualizing those functions or converging them with routers and switches; and kill the momentum of upstart Arista Networks in high-frequency financial trading and low latency switching environments.
The first three are summed up nicely in this Seeking Alpha post by consultant and financial analyst Nikos Theodosopoulos. He and we expect Cisco to unveil the Nexus 9000 line, the Insieme Networks-developed hardware foundation of its ACI anti-SDN, November 6 in New York.
[PRACTICAL EXPERIENCE: IT shops share OpenFlow, SDN best parctices]
Cisco has already expounded on the drawbacks of SDNs and the benefits of a physical/virtual hybrid like ACI. And the competitive stakes for Cisco have been well-documented as well.
But Theodosopoulos also explains that Cisco is coming at this whole SDN thing as VMware did in server virtualization years ago. VMware focused on server virtualization from the server, naturally; Cisco is basing network virtualization on the physical network.
Indeed, Cisco's basing its entire IT infrastructure, be it physical, virtual or both, on the network. Insieme's Nexus 9000 is expected to converge UCS servers and storage - recently-acquired WHIPTAIL's flash memory - with application aware switching and ACI control/management/orchestration in what Cisco is calling "the next generation of data center infrastructure."
In so doing, Cisco is looking to undercut VMware/Nicira's server hypervisor approach to network virtualization. And Theodosopoulos expects Nexus 9000 price points to also marginalize the commodity attraction of merchant silicon-based "white box" switches and other less expensive alternatives, like Arista. Arista has gained considerable traction in low latency switching, especially in the financial vertical, and is expected to soon file to go public.
Perhaps that's why the Nexus 9000 is apparently debuting with Broadcom silicon: to keep price points low, initially, until critical mass is achieved before the custom silicon-based ACI infrastructure comes along afterward.
While Cisco gets ready to redefine the data center, it is also setting its sights on the radio access network market, according to this post in istockanalyst.com. In addresses at Interop NY and Cisco Live, CEO John Chambers hinted that the company might be inclined to address the RAN market to complete its mobility portfolio. He also recently stated intentions to compete more directly with Ericsson and Nokia, two mobile infrastructure titans:
The logic has a rationale as the service provider-focused, licensed radio access piece is a key missing component of Cisco's much improved wireless arsenal.
The istockanalyst site values the RAN market at $40 billion.
Cisco currently offers RAN-to-IP gateways and optimization products. It has also acquired three companies with technologies pertinent to this space: Ubiquisys, BroadHop and Intucell.
The istockanalyst author, citing research from investment firm UBS, believes Cisco may employ the spin-in mode of market insertion for RAN, just as it did with Andiamo Systems, Nuova and most-recently Insieme. Cisco invests in an internally incubated start-up and then acquires it after product development milestones are achieved.
More from Cisco Subnet:Cisco Subnet bloggers on Twitter.Jim Duffy on Twitter