Broadcom has begun shipping its Tomahawk 6 networking chip, a high‑performance switch that delivers 102.4 terabits per second of switching capacity. The chip, which entered full production in less than three quarters from initial sampling, is now available to customers and is positioned to support the next generation of AI data‑center clusters.
The Tomahawk 6 doubles the throughput of its predecessor, Tomahawk 5, which offered 51.2 Tbps. In addition to the doubled bandwidth, the new switch features advanced SerDes configurations, including 512 × 200 G or 1024 × 100 G links, and supports co‑packaged optics (CPO) to reduce latency and power consumption in dense AI workloads.
Broadcom’s senior vice president and general manager of the Core Switching Group, Asad Khamisy, said, “Broadcom's ability to take Tomahawk 6 to production in less than three quarters from initial sampling is a testament to its innovation at scale. Tomahawk 6 doubles the throughput of its predecessor Tomahawk 5, representing a true breakthrough in AI infrastructure design. Our superior execution enables us to be the trusted technology partner for our customers' demanding networking needs.”
The switch is already shipping to customers, with early adopters deploying it in large AI clusters to achieve lower latency and higher power efficiency. While specific customer names are not disclosed, the product’s availability signals a shift toward higher‑bandwidth, lower‑latency fabrics in hyperscale data centers.
In the competitive landscape, the Tomahawk 6 positions Broadcom against NVIDIA’s InfiniBand solutions and other high‑performance networking options. By offering an open, standards‑based fabric that is Ultra Ethernet Consortium compliant, Broadcom provides a flexible alternative that avoids proprietary lock‑in while delivering the bandwidth required for AI workloads.
Ram Velaga, senior vice president and general manager of the Core Switching Group, added, “Tomahawk 6 is not just an upgrade – it's a breakthrough. It marks a turning point in AI infrastructure design, combining the highest bandwidth, power efficiency, and adaptive routing features for scale‑up and scale‑out networks into one platform. Demand from customers and partners has been unprecedented. Tomahawk 6 is poised to make a rapid and dramatic impact on the deployment of large AI clusters.”
The launch of Tomahawk 6 is part of Broadcom’s broader AI infrastructure strategy, which includes third‑generation CPO Ethernet switches, high‑bandwidth optical DSPs, and custom AI accelerators. The company’s focus on open standards and co‑packaged optics is intended to simplify deployment and reduce the optical complexity that has traditionally limited AI cluster scaling.
By delivering 102 Tbps of switching capacity, the Tomahawk 6 enables AI data‑center operators to scale clusters with fewer switch tiers, lower power draw, and reduced latency. This capability is expected to accelerate the adoption of large‑scale AI models and support the growing demand for high‑throughput, low‑latency networking in cloud and hyperscale environments.
Broadcom’s shipping of the Tomahawk 6 marks a significant milestone in the company’s AI portfolio, reinforcing its leadership in high‑performance networking and positioning it to capture a larger share of the rapidly expanding AI infrastructure market.
The content on EveryTicker is for informational purposes only and should not be construed as financial or investment advice. We are not financial advisors. Consult with a qualified professional before making any investment decisions. Any actions you take based on information from this site are solely at your own risk.