拥有广泛的产品资源,满足各种规模的计算需求

可提供智能、高效、敏捷、可靠的基础设施。

首页 > 产品中心 > 智能联网 > NVIDIA 网络 > NVIDIA 网卡 > NVIDIA InfiniBand 网卡 > NVIDIA MCX545B-ECAN ConnectX-5 IB网卡

NVIDIA MCX545B-ECAN ConnectX-5 VPI Adapter Card for OCP with Host Management EDR InfiniBand and 100GbE Single-Port QSFP28 PCIe 3.0 x16 No Bracket Type-1 Heat Sink

型号
NVIDIA MCX545B-ECAN ConnectX-5 IB网卡
编号
MCX545B-ECAN
价格
11500.00
热线
010-82117900
商家
北京慧点数码科技有限公司
保修
原厂全国联保
结算
银行汇款/转账支票/京东慧采

NVIDIA MCX545B-ECAN ConnectX-5 IB网卡

NVIDIA MCX545B-ECAN ConnectX-5 VPI Adapter Card for OCP with Host Management EDR InfiniBand and 100GbE Single-Port QSFP28 PCIe 3.0 x16 No Bracket Type-1 Heat Sink

ConnectX-5 VPI Adapter Card for Open Compute Project (OCP)
Intelligent RDMA-enabled network adapter card with advanced application offload and Multi-Host capabilities for High-Performance Computing, Web2.0, Cloud, and Storage platforms.
ConnectX-5 with Virtual Protocol Interconnect?? supports 100Gb/s InfiniBand and Ethernet connectivity, sub-600ns latency and very high message rate, plus PCIe switch and NVMe over Fabric offloads, providing the highest performance and most flexible solution for Open Compute Project servers and storage appliances while supporting the most demanding applications and markets: Machine Learning, Data Analytics, and more.

Features
    Tag matching and rendezvous offloads
    Adaptive routing on reliable transport
    Burst buffer offloads for background checkpointing
    NVMe over Fabric (NVMe-oF) offloads
    Back-end switch elimination by host chaining
    Enhanced vSwitch/vRouter offloads
    Flexible pipeline
    RoCE for overlay networks
    PCIe Gen 4 support
    RoHS compliant
    ODCC compatible

Benefits
    Up to 100Gb/s connectivity per port
    Open Compute Project form factor; OCP Specification 2.0 type 1 & type 2
    Industry-leading throughput, low latency & CPU utilization, and high message rate
    Multi-host technology maximizes data center ROI
    Innovative rack design for storage and Machine Learning
    Smart interconnect for x86, Power, Arm, and GPU-based compute and storage platforms
    Advanced storage capabilities including NVMe over Fabric offloads
    Support for flexible pipeline programmability
    Cutting-edge performance in virtualized networks including NFV
    Enables efficient service chaining
    Efficient I/O consolidation, lowering data center costs and complexity
* 与 NVIDIA 产品相关的图片或视频 (完整或部分) 的版权均归 NVIDIA Corporation 所有