Addressing AI’s Data Handling Challenges
Riverbed, a company specializing in network performance solutions, is making a significant move to address the escalating demands placed on network infrastructure by artificial intelligence workloads. The company has announced the rollout of its enhanced SteelHead lineup, featuring the new RiOS 10 acceleration software. This update represents Riverbed’s most comprehensive acceleration enhancement in seven years, designed to significantly boost the capabilities of its SteelHead hardware and software portfolio and tackle the complex data handling challenges that AI brings to modern networks.
Timely Enhancements for Reconfigured Networks
The timing of Riverbed’s announcement appears particularly relevant, coinciding with a period where enterprise networks are increasingly being reconfigured and optimized specifically to support the unique requirements of AI workloads. The enhancements introduced in the SteelHead products are aimed at empowering IT teams with the necessary tools to efficiently manage the now-massive data flows that AI tools and applications require for effective operation.
Speed Meets Security Across Diverse Environments
A core promise of Riverbed’s updated SteelHead lineup with the new RiOS 10 acceleration software is the ability to help companies move data faster across their distributed IT environments. This includes accelerating data transfer between various locations such as clouds, traditional data centers, and edge computing sites. Crucially, this speed enhancement is intended to be achieved without sacrificing essential network security or operational flexibility. Alongside these significant updates to its hardware and software components, Riverbed is also introducing a more adaptable approach to licensing its solutions, aiming to provide customers with greater flexibility in how they consume the technology.
CEO Highlights Scale of Update and AI’s Impact
According to David Donatelli, CEO of Riverbed, the rapidly changing networking landscape, driven by the accelerating adoption of AI, was a principal factor behind the company’s latest strategic developments. Donatelli stated that this dynamic environment “pushed Riverbed to reimagine how data moves through data centers, clouds, and edge environments.
” He elaborated on the significance of this release, noting, “This is our largest launch in seven years, principally driven by the new ways people need to configure their networks in this new world of AI, with a combination of new software, faster hardware appliances, and flexible business practices that allow our customers to consume our solutions in the ways they want.” The underlying message from Riverbed is clear: as AI continues to transform networking requirements, businesses are in need of what the company terms “fresh and ferocious approaches” to effectively handle their evolving data patterns and infrastructure needs.
Industry Analyst Views on AI and Networking
Industry analysis supports the notion that AI is fundamentally changing networking requirements. Jim Frey, a principal analyst focusing on networks at Enterprise Strategy Group, commented on this trend, stating, “AI has made networking much more critical again, across all different parts of the network.” Frey highlighted specific findings from their research, noting, “The research we did has pretty clear findings. AI drives the need to add bandwidth capacity, and enterprises need better, lower latencies because of a variety of ways that AI gets used.”
Generative AI’s Unique Data Challenges
Frey further explained the specific challenges posed by AI, particularly generative AI. These systems constantly collect massive amounts of data, both large and small, primarily for the purpose of training their underlying language models. This necessitates companies gathering significantly more data than they may have in the past and efficiently shipping it back to their centralized or distributed training systems. Beyond the training phase, generative AI presents another challenge during the inference or prediction stage.
When these systems generate predictions or responses, they often need to pull information rapidly from a central data location to wherever the AI computation is being performed. This creates what Frey describes as a “double whammy” for networks—not only is there a requirement for increased bandwidth to move larger volumes of data, but any delays or high latencies in this process can significantly hurt the performance and responsiveness of the AI applications.
Doubling Performance, Halving Costs
Riverbed’s latest upgrades are presented as representing a major performance leap, particularly for organizations already utilizing or considering their solutions. The updates reportedly double the throughput capacity of their appliances, increasing it from 30 Gbps to a substantial 60 Gbps per appliance. This enhanced capacity is designed to directly address the increased bandwidth requirements driven by AI workloads.
Substantial Cost Benefits Highlighted
In addition to the performance improvements, the cost benefits associated with Riverbed’s updated solutions are highlighted as substantial. Chalan Aras, who heads up the Acceleration division at Riverbed as Senior Vice President and General Manager, pointed out that their solutions now enable companies to move data at approximately half the cost of traditional methods. This significant price advantage, combined with the doubled performance capacity, positions Riverbed’s technology as an increasingly attractive option for companies grappling with the rapidly expanding data requirements and complex network demands imposed by the widespread adoption of artificial intelligence across their operations.
