Overview
Leveraging faster speeds and innovative In-Network Computing, NVIDIA ConnectX smart adapters achieve extreme performance and scale. NVIDIA ConnectX lowers cost per operation, increasing ROI for high-performance computing (HPC), machine learning, advanced storage, clustered databases, low-latency embedded I/O applications, and more.
  • Industry-leading throughput and latency performance
  • I/O consolidation
  • Virtualization acceleration
  • TCP/UDP/IP and iSCSI stacks
  • Dual Gigabit Ethernet ports
  • PCI Express 3.0
  • Low latency RDMA over Ethernet
  • Data center bridging support
  • T11.3 FC-BB-5 FCoE
  • TCP/IP stateless offload in hardware
  • Traffic steering across multiple cores
  • Hardware-based I/O virtualization
  • Intelligent interrupt coalescence
  • Advanced Quality of Service
  • RoHS-R6
World-class Ethernet performance RDMA over Converged Ethernet - ConnectX-3 utilizing IBTA RoCE technology provides efficient RDMA services, delivering low-latency and high-performance to bandwidth and latency sensitive applications. With link-level interoperability in existing Ethernet infrastructure, network administrators can leverage existing data center fabric management solutions. Sockets acceleration Applications utilizing TCP/UDP/IP transport can achieve industry-leading throughput over 10 or 40 GbE. The hardware-based stateless offload and flow steering engines in ConnectX-3 reduce the CPU overhead of IP packet transport, freeing more processor cycles to work on the application. I/O virtualization ConnectX-3 EN with Virtual Intelligent Queuing (Virtual-IQ) technology provides dedicated adapter resources and guaranteed isolation and protection for virtual machines (VM) within the server. ConnectX-3 EN gives data center managers better server utilization and LAN and SAN unification while reducing costs, power, and complexity. Precision data centers ConnectX-3 EN IEEE 1588 precision time protocol circuitry synchronizes the host clock to the data center master clock for accurate data delivery time stamping and data center SLA measurements. The hardware-based mechanisms ensure high accuracy and low jitter. Storage acceleration A consolidated compute and storage network achieves significant cost-performance advantages over multi-fabric networks. Standard block and file access protocols can leverage RDMA for high-performance storage access. T11 compliant encapsulation (FCoE) with full hardware offloads simplifies the storage network while keeping existing Fiber Channel targets. Quality of Service Resource allocation per application or per VM is provided and protected by the advanced QoS supported by ConnectX-3 EN. Service levels for multiple traffic types can be based on IETF DiffServ or IEEE 802.1 p/Q allowing system administrators to prioritize traffic by application, virtual machine, or protocol. This powerful combination of QoS and prioritization provides the ultimate fine-grained control of traffic - ensuring that applications run smoothly in today's complex environments.