VMICPCI-0569 | 反射式存储器接口 | 具有快速、准实时的特点

型号：VMICPCI-0569

类别：反射式存储器接口

特征：

网络传输速率 43 MB/s（4 字节数据包）至 174 MB/s（64 字节数据包）

动态数据包大小，4 到 64 字节的数据

多模光纤连接距离可达 300 米，单模光纤可达 10 公里

多达 256 个节点

GE Fanuc / VMICPCI-0569 反射式存储器接口允许在多达256个独立系统（节点）之间以高达174 MB/s的速率共享数据。每个反射式存储器板可配置 64 MB 至 256 MB 的板载 SDRAM。本地 SDRAM 提供对存储数据的快速读取访问时间。写入存储在本地 SDRAM 中，并通过高速光纤数据路径广播到其他反射式存储器节点。节点之间的数据传输是软件透明的，因此不需要 I/O 开销。发送和接收FIFO在峰值数据速率下缓冲数据，以优化主机和总线性能，以保持高数据吞吐量。

VMICPCI-0569 反射式存储器接口，通常指的是反射内存（Reflective Memory）的接口。反射内存具有快速、准实时的特点，主要依赖于光纤传输，用于分布式系统，使得多机能够协同完成某项任务或基于一个模型开展不同的运算和仿真分析。

VMICPCI-0569 反射内存实时网络提供一个所有节点同时共享的内存区域，且每个节点拥有该内存区域的物理反射。写入某个节点反射内存空间的数据，将被同时发送到其他所有节点。反射内存实时网络把来自不同节点的数据合并为统一的数据，并更新所有节点。这种设计保证了数据能够同时到达所有节点，并保持一致的字节顺序。

Model: VMICPCI-0569

Category: Reflective memory interface

Features:

Network transfer rate 43 MB/s (4-byte packets) to 174 MB/s (64-byte packets)

Dynamic packet size, 4 to 64 bytes of data

Multi-mode fiber connections can be up to 300 meters, and single-mode fiber connections can be up to 10 kilometers

Up to 256 nodes

The GE Fanuc/VMICPCI-0569 reflective memory interface allows data to be shared between up to 256 independent systems (nodes) at rates of up to 174 MB/s. Each reflective memory board can be configured with 64 MB to 256 MB of on-board SDRAM. Local SDRAM provides fast read access time to stored data. Writes are stored in local SDRAM and broadcast to other reflective memory nodes via high-speed fiber data paths. Data transfers between nodes are software-transparent, so there is no I/O overhead. Transmit and receive FIFOs buffer data at peak data rates to optimize host and bus performance to maintain high data throughput.

VMICPCI-0569 Reflective Memory interface, usually refers to the interface of reflective memory. Reflected memory is fast and quasi-real-time, mainly relying on fiber optic transmission, and is used in distributed systems, so that multiple machines can cooperate to complete a certain task or carry out different calculations and simulation analysis based on a model.

The VMICPCI-0569 Reflected Memory real-time network provides a memory region that is shared by all nodes simultaneously, and each node has a physical reflection of that memory region. Data written to the reflected memory space of one node is simultaneously sent to all other nodes. The reflected memory real-time network combines data from different nodes into a unified data and updates all nodes. This design ensures that the data reaches all nodes at the same time and maintains a consistent byte order.