LAS VEGAS – Cisco emphasized the role of AI at its Live customer event this week, highlighting new network, management, and security solutions, as well as forming partnerships and investments aimed at fostering the use of AI in businesses.
“AI is evolving at a phenomenal pace, much faster than many technologies we’ve seen in the past. Consider how long we discussed 5G before it was actually implemented,” explained Cisco CEO Chuck Robbins during his keynote. “Generative AI is progressing rapidly, and it’s being integrated throughout our product lineup. You’re likely exploring potential applications, if you haven’t already started, and how they might benefit your operations.”
Robbins described how the future of hyper-extended enterprises could be vastly simplified by the ability to monitor the entire technology stack and infrastructure, right down to the network level, facilitating better data utilization to predict and avert system downtimes and other issues. He likened the rapid adoption of AI to the cloud revolution, but intensified.
Cisco’s aggressive AI initiatives involve the establishment of networked AI clusters and leveraging AI for efficiently managing networking environments.
On the networking front, Cisco introduced a complete AI solution named Cisco Nexus HyperFabric AI cluster, expected to be tested in the fourth quarter. This solution allows customers to set up an integrated networking system using a Cisco 6000 series switch for spine and leaf setups that support 400G and 800G Ethernet fabrics, GPUs. This is in a partnership with Nvidia, including Nvidia’s BlueField-3 data processing unit (DPU) and SuperNIC, AI pod-building reference designs, and the Vast Data Platform for unified storage, databases, and a data-driven function engine designed for AI.
“HyperFabric is an automated, intelligent AI-native solution, designed to streamline and introduce AI to the on-premise enterprise data center,” stated Jonathan Davidson, executive vice president and general manager of Cisco Networking. “This package eases the design, deployment, and management of networking, computing, and storage components required to build comprehensive AI solutions where enterprise data resides.”
On the management aspect, Cisco announced the launch of Digital Experience Assurance, which enhances its ThousandEyes network intelligence platform across cloud platforms, initially supporting Amazon Web Services. It enables customers to identify issues and failures within both the resources they manage and external ones, such as those provided by a cloud service. Cisco notes that the in-depth intelligence is driven by tens of thousands of ThousandEyes agents spread throughout the internet and enterprise networks, with more than 650 billion global measurements taken daily.
The company has also expanded its AI-driven cloud insights initiative. Once customer workloads are hosted on a cloud provider’s network, the tool provides a detailed view of how network elements and services are interconnected, how application load balancers tie into network load balancers, and how these relationships collectively form a target group spanning several AWS EC2 instances, explained Joe Vaccaro, vice president and general manager of Cisco ThousandEyes.
AI-driven intelligence in the ThousandEyes package can identify complex patterns and pinpoint issues, enabling prompt prioritization of incidents, according to Vaccaro. He emphasized the efficiency of ThousandEyes Event Detection in reducing time-consuming troubleshooting processes to a matter of minutes.
The recent enhancements to ThousandEyes Endpoint Experience incorporate the collection of data from Meraki Wi-Fi and LAN, providing deeper insights into local network disruptions that impact user experiences, as detailed by Vaccaro.
Cisco has integrated support for AMD Pensando DPUs within its latest AI-infused offering, HyperShield. This advanced security fabric aims to safeguard distributed applications and data with continuous updates. Pensando DPUs are notable for their software programmability across various IT domains and will be integrated into Cisco’s UCS servers along with offerings from other prominent server providers by the end of 2024. Additionally, Cisco intends to extend compatibility to Intel’s programmable infrastructure processing units (IPU).
Hypershassets utilizes AI-driven software, virtual environments, and contemporary technologies that will eventually integrate into fundamental networking frameworks like switches, routers, or servers. It is engineered to allow enterprises to independently compartmentalize their networks during security threats, enable swift exploit defenses without the need for patches or firewall modifications, and facilitate the seamless update of software without disrupting computing operations.
An innovative AI-focused management architecture named Security Cloud Control is slated for release. Its purpose is to configure, manage, and supervise the entire framework of Cisco’s Security Cloud, starting with the network security solutions offered by Cisco.
“Security Cloud Control is set to be the sole management interface for all of Cisco’s Security Cloud, streamlining operations for security administrators and IT professionals. We anticipate its initial release in October, supporting our network security offerings initially, with further expansions planned thereafter,” stated Jeetu Patel, the Executive Vice President and General Manager of Cisco Security & Collaboration. Patel emphasized that this development would simplify the configuration, management, and monitoring processes of Cisco’s cloud solutions for customers.
“Expanding past AI assistants, Security Cloud Control introduces an AI-native strategy that proactively provides actionable insights and automates solutions across hybrid ecosystems,” remarked Patel.
Cisco has introduced the Secure Firewall 1200 Series in its Security Cloud, which integrates sophisticated security capabilities with SD-WAN connectivity into a single device, according to Patel.
Furthermore, Cisco has released a new iteration of its firewall software, Firewall Threat Defense (FTD) version 7.6, which is available for all physical and virtual Cisco firewalls. FTD 7.6 employs AI to combat zero-day threats and has expanded application control to more than 70 generative AI applications to better protect sensitive data. It also enhances the efficiency of branch network deployments using pre-configured SD-WAN and firewall templates and supports zero-touch provisioning, as explained by Patel.
“We believe security is crucial for AI, and equally, AI is vital for enhancing security,” stated Robbins. “Effective security measures are essential to protect the data utilized in training AI models within private enterprises. It’s imperative to safeguard these models from cyber threats and ensure the queries being made do not conflict with organizational values or interests. Thus, maintaining robust security for AI continues to be a focus for us.”
Beyond product enhancements, Cisco has also detailed plans to advance AI-specific career training initiatives. The company has launched the preliminary phase of its AI partner specialization training, aimed at equipping business partners with the necessary skills to efficiently manage AI workloads on Cisco infrastructure. This training covers foundational AI concepts, a classification of AI solutions, and insights into Cisco’s AI strategy, governance, and relevant application cases on Cisco platforms.
Cisco has introduced a new AI certification focused on designing AI architecture. This vendor-neutral certification aims to equip both current and prospective IT professionals with essential skills for developing AI/ML compute and network architectures, as stated by Cisco.
In addition, Cisco is establishing a $1 billion global AI investment fund to support the startup community and enhance the creation of secure and dependable AI technologies. Within this initiative, Cisco has made investments in Cohere, Mistral AI, and Scale AI. Over recent years, Cisco has expanded its AI portfolio through acquisitions or investments in over 20 AI-centric companies.
The success of Cisco’s AI strategy is expected to significantly depend on the integration of multi-product solutions and strategic partnerships, according to industry experts.
Vijay Bhagavath, research vice president at IDC, highlighted the Nexus HyperFabric AI cluster as a noteworthy innovation. He noted its unique position in the market and its appeal to enterprise customers seeking to implement AI solutions efficiently without the necessity of sourcing multiple separate components.
In terms of its partner ecosystem, Cisco can afford to buy or support the companies it needs to build out its AI strategy and help the overall marketplace develop, Bhagavath mentioned. “It has a significant advantage in AI marketing but the landscape is still evolving, making it fascinating to see how it will unfold.”
Some experts believe Cisco should present a more unified AI strategy rather than releasing isolated solutions.
“Cisco needs to approach the AI market with a clear and strategic plan, rather than a fragmented approach,” stated Jim Frey, principal analyst of networking for the Enterprise Strategy Group. “At times, they seem to lag behind other AI players like Microsoft, but in truth, they are proceeding in a very thoughtful and systematic manner.”
Nonetheless, it’s still the beginning stages of AI, and Cisco is not unique in its quest to develop a comprehensive strategy for AI. Many competitors are attempting to do the same, according to Frey.
