At VMware Explore Barcelona, VeloCloud Rolls Out Enhancements to Scale AI

AI has moved from proof of concepts and kicking the tires to reality for most companies. While there are still many questions regarding the hows and whats of how to use AI, from a network perspective, the impact of AI is massive. AI will have as significant a transformative an effect on the network as the cloud did, particularly the WAN.

One supporting data point comes from a recent ZK Research study, which asked, “What are your organization’s biggest network challenges today?” I fully expected the top response to be “Ensuring appropriate levels of network security,” but that was surprisingly number two. The top response? Supporting AI workloads shows it having an impact today, and I predict that unless networks evolve, the problem will get bigger.

Requirements to Scale AI Grow

As more companies increasingly rely on AI, networks must handle larger amounts of data, keep performance steady, and ensure critical applications work smoothly. To help companies manage these growing requirements, Broadcom announced enhancements to its VMware VeloCloud product line by launching a robust AI networking architecture, new appliances, and a partner program at its recent Explore event in Barcelona.

First in the lineup is VeloRAIN (Robust AI Networking), a networking architecture designed to boost network performance and security for distributed AI apps. VeloRAIN utilizes AI and machine learning (ML) to recognize encrypted application traffic, which was previously difficult to analyze. This allows organizations to identify and prioritize new AI apps at the edge of their networks, improving service quality.

Additionally, VeloRAIN includes tools to optimize network efficiency, especially over wireless links like 5G and satellite. Whether these wireless links are used alone or together with wired connections, VeloRAIN helps maintain stable, high-quality performance, similar to fiber networks, even under changing conditions. Using this feature, organizations can speed up network setup in remote or branch offices while reducing the time and resources needed for deployment.

Another core component of VeloRAIN is dynamic application-based slicing (DABS), which automatically assigns priorities to different apps across various networks. DABS ensures that critical tasks get the attention they require. It can even prioritize traffic for specific users based on their roles and needs, enhancing the overall user experience. This feature ensures AI traffic can be prioritized ahead of other data to ensure optimal performance.

Addressing the Edge

In addition to VeloRAIN, Broadcom introduced two new high-performance edge appliances: the VeloCloud Edge 4100 and 5100. Tailored for large environments like regional hubs and data centers, these devices combine software-defined wide area networking (SD-WAN), security, and AI capabilities. The VeloCloud Edge 4100 handles up to 30 gigabits per second (Gbps) throughput and supports 12,000 connections, a significant leap from earlier models.

For even greater demand, the Edge 5100 supports up to 100 Gbps throughput and 20,000 connections, making it ideal for enterprises that require ultra-high performance. By reducing the number of devices, these appliances simplify the network architecture, offer greater resilience, and reduce latency. The technology addresses complex enterprise needs by scaling beyond 100 Gbps if necessary and supporting AI and non-AI workloads.

Lastly, Broadcom launched the Titan program for VeloCloud partners. Replacing the former VMware VeloCloud Partner Program, Titan will support managed service providers (MSPs) with a three-tier structure—pinnacle, premier, and registered—each offering tailored benefits to help partners expand their businesses. The benefits include performance-based incentives, opportunities for joint solution development, market visibility, and business growth.

Titan also features a new model that simplifies license management, offering greater flexibility and stable pricing. Through the program, partners will provide VeloCloud network services to enterprise customers globally as a managed service, while customers will benefit from the partners’ expertise.

According to findings from Broadcom’s new State of the Edge report, AI is increasingly driving productivity and reducing operational costs across industries. Most manufacturing organizations (93 percent) see AI as a catalyst for growth and innovation, while 80 percent of retail companies plan to implement AI automation by 2025. In telecommunications, generative AI is expected to fuel growth for early adopters.

The enhanced VeloCloud portfolio aims to support companies in managing complex and distributed workloads. These updates align with Broadcom’s goal of helping organizations leverage AI to modernize their operations and enhance efficiency.

Zeus Kerravala is the founder and principal analyst with ZK Research. Read his other Network Computing articles here.