Specifications
Ordering Code
E8102CQDA2 / E810-2CQDA2
Controller
2 × Intel Ethernet Controller E810
Ports
2 × 100GbE QSFP28 (or 50/25/10G)
Form Factor
PCIe add-in card (low profile / full height bracket options)
Host Interface
PCIe (see per-model lane/Gen detail above)
Bracket Options
Low-profile and full-height brackets supplied
Network Standards
IEEE 802.3 Ethernet; IEEE 802.1Q VLAN; IEEE 1588 PTP (model dependent)
Virtualization
SR-IOV, VMDq, MAC/VLAN filtering, 802.1Qbg edge virtual bridging
Offloads
TCP/UDP/IP checksum, TSO/LRO, RSS, DCB, FCoE/iSCSI (model dependent)
Management
PXE boot, iSCSI boot, UEFI option ROM, NC-SI (model dependent)
Operating Temperature
0 °C to +55 °C (server-class airflow required)
Compliance
RoHS, CE, FCC Class A
Source
Factory-sealed OEM / retail pack, traceable batch
Lead Time
In stock; 3-7 day dispatch for standard volumes
MOQ
1 unit (sample); volume tiers from 10 units
PCIe Host Interface
PCIe 4.0 x16 (16 GT/s)
Aggregate Bandwidth
200 Gb/s (2 × 100G independent)
Breakout
Each port 100G → 4 × 25G with breakout cable
Queue Pairs
Up to 256 per controller
RDMA
RoCEv2 and iWARP
Target Market
Storage heads, spine leafs, dual-rail AI fabric clients
Overview
The Intel E810-2CQDA2 carries two independent E810 controllers on a single PCIe 4.0 x16 adapter, giving a node two fully separate 100GbE rails — the standard dual-rail layout for AI training clients and high-availability storage heads.
Because the two ports are independent controllers, each has its own queue set and RDMA engine; a rail failure or a driver reset on one port does not interrupt traffic on the other, which is exactly the property multi-rail fabrics rely on.
Key Benefits
Two independent 100G rails from one slot. Dual E810 controllers for fault isolation. 200 Gb/s aggregate bandwidth. RoCEv2 + hardware PTP on both rails.
Applications
Dual-rail AI training and inference clients, high-availability NVMe-oF storage heads, spine/leaf aggregation nodes, and HPC message-passing fabrics.
Request a Quote — INTEL E810-2CQDA2 — DUAL-PORT 100GBE QSFP28 PCIE 4.0 X16 SERVER ADAPTER
QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.
Get Your Quote →