Intel Ethernet Network Adapter E830-CQDA2 — Dual-Port 200GbE OCP 3.0

Specifications

Overview

The Intel Ethernet Network Adapter E830-CQDA2 is a dual-port 200GbE adapter in an OCP 3.0 mezzanine form factor, built on Intel's E830 controller. It presents two 100GbE ports by default or a single 200GbE port, and each cage can be broken out to four 25GbE links — up to eight 25GbE ports or eight 10GbE ports from one card. Host attachment is PCIe 5.0 x8, with a PCIe 4.0 x16 option for platforms that cannot supply Gen5 lanes.

The E830 generation is the security-focused successor to the E810. Signed firmware, secure boot and a hardware root of trust are standard, and the adapter is CNSA 1.0 compliant — the requirements set that government and defence buyers use to screen network hardware. On the timing side it supports IEEE 1588 PTP v2 and Precision Time Measurement v1.0a, which matters for telecom, financial and industrial networks where timestamp accuracy below a microsecond is required.

Power is notably low for 200GbE: 11.8 W typical and 13.8 W maximum in the 2 × 100Gb traffic case, and 10.0 W typical for a single 200GbE port. Dynamic Device Personalization lets the packet-processing pipeline be re-profiled per workload, and DPDK support keeps it usable in NFV and packet-broker roles. QS Compute supplies Intel E830 adapters with OCP 3.0 and standard PCIe brackets.

Key Benefits

200GbE on an OCP 3.0 mezzanine — no full-height slot consumed, leaving PCIe for accelerators. CNSA 1.0 + hardware root of trust — a defensible security baseline for government and regulated buyers. Low power for the class — 11.8 W typical at 2 × 100Gb, 10 W at a single 200Gb port.

Applications

OCP 3.0 hyperscale and cloud servers, telecom and 5G core nodes requiring IEEE 1588 timing, storage and SDS front-end networking, security appliances, and NFV / packet-broker deployments using DPDK.

Related Products