Edgecore Networks, a global leader in open networking solutions, today announced the launch of its 400G-optimized spine switch, the DCS511, The new switch is designed to meet the demanding requirements of data center and cloud computing environments, offering 12.8 Terabits per second switching capacity and low-latency, making it ideal for data centers, service providers, and cloud operators.
The DCS511 is an open network data center switch powered by the Broadcom Tomahawk 4 chipset, offering 32x400G ports on a single device. It can be deployed as a spine switch supporting 100/400 GbE spine-to-spine or spine-to-leaf interconnects.
This open network switch is equipped with the Open Network Install Environment (ONIE), allowing the installation of compatible Network Operating System (NOS) software, including the open-source Open Network Linux, as well as commercial NOS offerings.
Edgecore’s DCS511 32-port Tomahawk4 switch offers a cost-effective solution for lower-capacity data centers seeking 400G options, providing an alternative to 32x400G Tomahawk3 Spine switch. The new DCS511 fabric offers Switching solution with low density 400G ports for General & AI Compute Datacenter workloads. The DCS511 offers the benefits of less port density, allowing organizations to tailor their network infrastructure to their specific needs, avoid over-provisioning, and reduce unnecessary costs.
Edgecore is committed to exploring the possibilities of data center architectures, delivering future-ready solutions for your AI/ML connectivity needs. Get ready to embrace a faster, high-capacity network with Edgecore’s DCS511 – high-capacity-network switch.
Key Features of the DCS511:
- Up to 16 ports for 400G ZR or ZR+ coherent optics, with a 24-watt power budget per port
- Supports SONiC NOS for webscale and enterprise grade datacenters
- Pre-loaded with the Open Network Install Environment (ONIE) for automated loading of compatible open-source and commercial NOS offerings
- Supports VXLAN routing and bridging compared to TH3 switch
- Supports cognitive routing / GLB / path rebalancing / RoCEv2 over VXLAN feature for AI and LLM (Larger Language Model)