Analyst(s): Ron Westfall
Publication Date: November 4, 2024
Cisco unveils Plug-and-Play use case and industry-specific AI PODs aimed at making it easier for partners to sell and customers to deploy AI infrastructure
What is Covered in this Article:
- Cisco expands UCS portfolio with an AI server family for NVIDIA accelerated computing.
- New solutions with NVIDIA are designed to be highly adaptive and scalable for the evolving demands of enterprises.
- With an upgraded portfolio, powered by new 800G switches, Cisco spotlights strategy to accelerate enterprise AI adoption by offering open, scalable, and easier paths to AI infrastructure.
The News: Cisco announced new additions to its data center infrastructure: an AI server family purpose-built for GPU-intensive AI workloads with NVIDIA accelerated computing, and AI PODs to simplify and de-risk AI infrastructure investment. These are designed to give organizations an adaptable and scalable path to AI, supported by Cisco’s networking capabilities.
The Cisco UCS C885A M8 Server is now orderable and is expected to ship to customers by the end of this year. The Cisco AI PODs will be orderable in November. The Cisco Nexus 9364E-SG2 Switch will be orderable in January 2025 with the availability to begin Q1 calendar year 2025. Cisco Nexus Hyperfabric will be available for purchase in January 2025 with 30+ certified partners. Hyperfabric AI will be available in May and will include a plug-and-play AI solution inclusive of Cisco UCS servers along with embedded NVIDIA accelerated computing and AI software, and optional VAST storage.
Cisco Ups the AI Ante Launching New Plug-and-Play Solutions with NVIDIA
Analyst Take: Cisco’s latest offerings equip customers with essential infrastructure components to enhance their AI adoption, regardless of their current capabilities. These advancements build upon existing systems, allowing organizations to expand and innovate while minimizing complexity. Managed through Cisco Intersight, these solutions facilitate centralized control and automation, streamlining processes from configuration to daily operations.
Cisco is expanding its UCS AI compute lineup with the introduction of the UCS C885A M8 servers, specifically engineered for AI workloads that require intensive GPU usage. These powerful, high-density servers are designed to handle the most challenging AI training and inference tasks, leveraging the capabilities of NVIDIA’s HGX supercomputing platform equipped with NVIDIA H100 and H200 Tensor Core GPUs.
To enhance AI networking performance, each server is outfitted with NVIDIA NICs or SuperNICs, while NVIDIA BlueField-3 DPUs are incorporated to boost GPU data access and implement robust, zero-trust security measures. This release marks Cisco’s inaugural venture into the dedicated AI server market and represents the company’s first eight-way accelerated computing system constructed on the NVIDIA HGX platform.
Moreover, Cisco’s new AI PODs offer a comprehensive approach for organizations looking to implement AI infrastructure. These pre-configured stacks combine compute, networking, storage, and cloud management components, optimized for specific AI applications and industries. Based on Cisco Validated Designs, the AI PODs provide a reliable foundation that can be customized to meet unique requirements. AI PODs can simplify the deployment of AI inference solutions according to customer scaling needs, from edge computing to large-scale clusters powered by NVIDIA accelerators.
By incorporating NVIDIA AI Enterprise software, these solutions can accelerate data science workflows and streamline AI development and deployment processes. This integrated approach enables faster implementation, consistent performance, and reduced risk for AI initiatives, with the end goal of allowing organizations to swiftly realize value from their AI investments.
Cisco Focuses on Accelerating AI Compute
The journey into AI typically begins with the crucial phase of training generative AI (GenAI) models using vast amounts of data to develop model intelligence. Cisco’s new UCS C885A M8 Server is specifically engineered to meet this task head-on. The server is built to handle the most demanding AI training tasks, due in good part to its high-density configuration featuring NVIDIA H100 and H200 Tensor Core GPUs.
The server’s performance is further enhanced by the NVIDIA HGX architecture and AMD EPYC processors, providing the computational power required to process massive datasets and complex algorithms. In my view, what can set the UCS C885A M8 apart is its raw power, alongside its user-friendly design, offering simplified deployment and streamlined management. This combination of power and ease of use can make it more compelling for enterprise customers looking to dive into AI without getting bogged down by technical complexities.
The training of GenAI models requires synchronized operation of powerful server clusters, generating massive data flows that demand a high-performance network fabric. The Cisco Nexus 9364E-SG2 Switch is developed to meet these requirements as its 800G high-density aggregation capabilities can deliver efficient data transfer between servers, while sophisticated congestion management and ample buffer capacity minimize packet loss. As a potential key component of scalable network infrastructure, the Nexus 9364E-SG2 supports expansion of AI clusters to accommodate growing demands.
Cisco Shrewdly Targets Fast-Growing Global Data Center GPU Market
Cisco is making solid bets in broadening its targeting of the AI infrastructure market. Futurum Intelligence data indicates the global data center GPU market is set to experience rapid expansion, driven by increasing demand for AI and machine learning applications, cloud-based services, and evolving competitive dynamics. This is a major impetus for greater investment in AI infrastructure across hybrid AI and hybrid cloud environments.
According to Futurum Intelligence data, 2023-24 global AI deployment shows a strong preference for hybrid environments, with $16.1 billion invested in hybrid solutions. Cloud deployments follow, accounting for $8.04 billion in AI spending, indicating the growing importance of cloud-based infrastructure. On-premises deployments represent $3.6 billion, reflecting a smaller but still significant role in sectors requiring enhanced control and security.
Moreover, based on forecasts from Futurum Intelligence, GPU revenue is projected to grow at a 29.9% CAGR over the next five years, reaching $103 billion by 2028. Starting from $27.7 billion in 2023, revenue is expected to rise to $41.5 billion in 2024 (49.6% growth), $54.7 billion in 2025 (31.8%), and $72.7 billion in 2026 (33.0%), eventually slowing to 16.2% by 2028.
Smarter Decision-Making with AI-Powered Analytics and Capabilities
From my point of view, one of the primary benefits of integrating AI into business operations is the ability to derive actionable insights from data. AI models, particularly those driven by machine learning and deep learning, can identify patterns, trends, and anomalies within data that would be difficult or impossible for humans to detect manually.
I find that a key advantage here is speed and accuracy. Traditional data analysis methods are often slow and resource-intensive, which can delay decision-making. In contrast, AI models running on NVIDIA GPUs can process vast amounts of data in parallel, delivering insights faster and with greater accuracy. This is critical for industries where timely decision-making is essential, such as in warehouses, where faster intelligence can lead to better supply chain outcomes, or in logistics, where optimized routing can reduce costs and improve delivery times.
Also, in manufacturing, AI can be used to monitor equipment in real time, predicting failures before they happen and scheduling maintenance accordingly. This predictive maintenance capability could significantly reduce downtime and improve the overall efficiency of operations. Similarly, in customer service, AI-powered chatbots and virtual assistants can handle routine inquiries, freeing up human agents to focus on more complex tasks.
Cisco Ready to Alter Competitive Dynamics of the AI Infrastructure Market
The move strengthens Cisco’s competitive stance against AI infrastructure rivals. As I noted, the UCS C885A M8 represents the company’s initial foray into the dedicated AI server market alongside its debut of an eight-way accelerated computing system built on the NVIDIA HGX platform.
The new Cisco UCS servers expand its portfolio to counter HPE’s pending post-Juniper acquisition portfolio and strategic position. Through the acquisition of Juniper, HPE looks to enhance its existing portfolio by making a calculated leap into driving high-growth production AI business opportunities using a turnkey full-stack private cloud AI offering that includes AI-powered network automation and advanced network security. The pending acquisition consolidates HPE’s position in the rapidly evolving AI infrastructure market, aligning fully with the industry’s shift toward AI workload optimization.
NVIDIA AI Computing by HPE proposition incorporates ProLiant DL servers to address customer compute requirements. For example, the HPE ProLiant Compute DL380a Gen12 is targeted at delivering AI fine tuning and inferencing for large workloads using NVIDIA GPU acceleration capabilities. Now Cisco has a more direct in-house counter to HPE ProLiant DL server products. Likewise, the Cisco Nexus 9364E-SG2 Switch expands the range of alternatives to Juniper’s QFX/EX switch families such as the QFX5200.
Although the debut of the new Cisco server and switch offerings follow on the existing HPE/Juniper equivalent offerings, Cisco is positioned to close near-term HPE/Juniper time to market advantages as HPE goes through the process of integrating Juniper into the overall organization, which is likely to commence around Q1 2025.
Cisco’s move also intensifies competition in relation to Dell and its PowerEdge AI server suite. Dell already debuted a total of 15 next-generation Dell PowerEdge systems that can draw from NVIDIA’s full AI stack—including GPUs, DPUs, and the NVIDIA AI Enterprise software suite—providing enterprises the foundation required for a wide range of AI applications, including speech recognition, cybersecurity, recommendation systems, and a growing number of language-based services.
Subsequently, Dell AI Factory is being developed as an end-to-end AI enterprise solution integrating Dell’s compute, storage, client device, software, and services capabilities with NVIDIA’s advanced AI infrastructure and software suite, all underpinned by a high-speed networking fabric. Specifically, Dell PowerEdge XE 9680 servers will now support new NVIDIA GPU models, including the NVIDIA B200 Tensor Core GPU, expected to offer up to 15 times higher AI inference performance and lower total cost of ownership. Dell PowerEdge servers will also support other NVIDIA Blackwell architecture-based GPUs as well as H200 Tensor Core GPUs and NVIDIA Quantum-2 InfiniBand and Spectrum-X Ethernet networking platforms.
Also, the Lenovo ThinkSystem SR680a V3, SR685a V3, and SR780a V3 GPU systems are also built to deliver massive computational performance for AI, HPC, and graphical & simulation workloads across various industries. The family of servers supports eight high-performance GPUs, either from AMD or NVIDIA, with planned support for GPUs from Intel.
However, at the Cisco Partner Summit 2024, the two companies unveiled offerings that build on their existing partnership to simplify work environments, optimize enterprise infrastructure, and clear the path to enterprise AI through leveraging aspects of Cisco’s Collaboration Devices, networking, and security. Among the solutions announced were Truscale Meeting Room as a Service (TS MRaaS), TruScale WorkSmart, Factory and Rack Integrated solutions, ThinkAgile HCI hybrid cloud, and SAP HANA solutions with Cisco networking integration.
As such, while there is competitive overlap between the new Cisco UCS offering the Lenovo ThinkSystem SR650a/685a/780a family, I anticipate that both companies will effectively coordinate their channel and sales strategies to meet customer AI server and infrastructure preferences and priorities.
Looking Forward
Overall, I believe the new Cisco directly addresses the increasing pressure enterprise customers are experiencing to implement AI workloads, particularly to take advantage of the transition to agentic workflows and AI’s ability to autonomously address top business and technical priorities. Cisco’s advancements, including AI PODs and GPU servers, are solidly positioned to enhance the security, compliance, and processing capabilities of workloads as customers progress through their AI journeys from inferencing to training.
What to Watch:
- Cisco now offers a competitive lineup of products and solutions for data and performance-intensive use cases such as large language model training, fine-tuning, and inferencing for GenAI that can streamline enterprise adoption of production-ready AI.
- Enterprises will increasingly prioritize evaluating and selecting AI infrastructure solutions that support the entire AI lifecycle, from building and training sophisticated models to widespread use for inferencing.
- By leveraging the portfolio-wide capabilities of Cisco’s networking, security, and observability solutions, while recognizing the distinct paths customers take toward AI, Cisco and its partners can move the needle in providing genuine business value to their clients.
Disclosure: The Futurum Group is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.
Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of The Futurum Group as a whole.
Other insights from The Futurum Group:
Cisco Q4 & FY 2024: Focus on AI, Security, and Cloud
Lenovo and Cisco: Let’s Ease the Path to AI Innovation
HPE Unveils NVIDIA AI Computing by HPE: Enterprise AI Ascends
Author Information
Ron is an experienced, customer-focused research expert and analyst, with over 20 years of experience in the digital and IT transformation markets, working with businesses to drive consistent revenue and sales growth.
He is a recognized authority at tracking the evolution of and identifying the key disruptive trends within the service enablement ecosystem, including a wide range of topics across software and services, infrastructure, 5G communications, Internet of Things (IoT), Artificial Intelligence (AI), analytics, security, cloud computing, revenue management, and regulatory issues.
Prior to his work with The Futurum Group, Ron worked with GlobalData Technology creating syndicated and custom research across a wide variety of technical fields. His work with Current Analysis focused on the broadband and service provider infrastructure markets.
Ron holds a Master of Arts in Public Policy from University of Nevada — Las Vegas and a Bachelor of Arts in political science/government from William and Mary.
