World’s highest-performance ethernet switch-on-chip now shipping in production at the heart of Arista 7060X4 Series
The new systems announced by Arista, providing 32 ports of 400GbE connectivity over OSFP or QSFP-DD optical interfaces in a compact 1U hardware form factor, are each built using a single 12.8 Terabit/sec Tomahawk 3 BCM56980 device to provide full-duplex, line rate switching and routing across all ports. The Tomahawk 3 Series provides industry-leading 50G PAM-4 SerDes performance, the highest available Ethernet I/O and packet processing bandwidth on a single chip at 100/200/400GbE, as well as state-of-the-art buffering, load balancing, high-availability and network telemetry capabilities, tailored to the rigorous demands of software-defined cloud networks. Combined with Arista’s robust, unified EOS® network operating system and optical integration of OSFP and QSFP-DD technologies, Tomahawk 3 enables the new 7060X4 systems to accelerate adoption of 400GbE network infrastructure in the cloud.
“Tomahawk 3 is an important catalyst for the build-out of 400 Gigabit Ethernet cloud networks,” said
StrataXGS Tomahawk 3 Series Product Highlights:
- Supports line-rate, non-blocking 32 x 400GbE switching and routing on a single chip, with fully flexible I/O configuration to enable breakout to 64 x 200GbE ports, 128 x 100GbE ports, or any combination thereof
- Delivers quantum leap in hyperscale network performance,
CAPEX and OPEX efficiency: 40% reduction in power per 100Gbps, and up to 75% lower cost per 100Gbps, versus alternatives - New, state-of-the-art, integrated shared-buffer architecture offers 3X to 5X higher incast absorption and provides the highest performance and lowest end-to-end 99.9% latency for RoCEv2 based workloads
- Industry-leading Broadview™ Gen 3 integrated network instrumentation feature set and software suite, providing full visibility to network operators into packet flow behavior, traffic management state, and switch internal performance
- Comprehensively supports all packet processing and traffic management requirements for cloud-scale network use cases: up to Million-route forwarding scale, doubled ECMP scale, Dynamic Load Balancing and Group Multipathing, In-Band Network Telemetry, Elephant Flow detection and re-prioritization
- Robust connectivity using 256 instances of the industry’s best performing and longest-reach 50G PAM-4 integrated SerDes core, enabling long-reach (LR) East-West optical links and Direct-Attached-Copper (DAC) in-rack cabling in the data center, fully compliant to new IEEE standards for 50/100/200/400GbE
- Multiple switch device variants at 12.8 Tbps (32x400GbE / 128x100GbE), 8.0 Tbps (48x100GbE + 8x400GbE), and 6.4 Tbps (16x400GbE / 64x100GbE), providing optimized performance, cost, and power based on application use case
- Implemented with unparalleled silicon area and power efficiency; qualified and shipping in volume on proven 16nm process technology node
About
Press Contact:
Corporate Communications
david.szabados@broadcom.com
Telephone: 1-408-433-7848