參數(shù)資料
型號: IBM3288H2848
英文描述: Telecommunication IC
中文描述: 通信集成電路
文件頁數(shù): 20/131頁
文件大小: 1679K
代理商: IBM3288H2848
PRS28.4G
IBM Packet Routing Switch
Architecture
Page 20 of 131
prs28.03.fm
August 31, 2000
2.1.11.2 Line Interface and Synchronization
Data are transmitted on the physical lines two bits at a time, at a rate four times that of the internal byte clock.
Each bit is carried by a differential pair at a speed of 444 Mb/s. Both the transmitting and receiving devices
run with the same byte clock.
Link synchronization is provided by a training sequence of special Idle Packets, called Sync Packets. When
sync packets are transmitted on a port, the receiving port can perform bit phase alignment of the incoming
data, as well as packet alignment. The format of these sync packets is such that they are recognized by the
receiving end as valid packets, and provide bit transitions on the physical line to allow for phase recovery. A
sync packet LU is entirely composed of ‘CC’ bytes, except for the last byte which is ‘33’.
One data byte is carried over two differential pairs. One pair carries all even numbered bits, and the other one
all odd numbered bits. This converts the ‘CC’ sync packet bytes into 0 and 1 bit transitions on the physical
lines.
2.1.11.3 Link Integrity
Link integrity is provided by a CRC field, placed in the last byte of all Idle Packet LUs. The 8-bit CRC is the
checksum of all bits carried over a port since the previous Idle Packet ended. Thus, when multiple Data
Packets are transmitted, followed at some time by an Idle Packet, the CRC of this Idle Packet covers all
previous Data Packets. This provides a measure of the link quality for fault isolation.
2.1.11.4 Packet Numbering
In order to extract the output queue grant information that is multiplexed in the second and third bytes of
outgoing packets, Idle Packets carry a packet numbering field. This field is equal to the value of the priority of
the grant status being carried by the current Idle Packets.
The number of packet cycles required to carry the output queue grant for all priorities is equal to the number
of priorities. This output queue grant cycle repeats itself as long as non-sync packets are transmitted. The
output queue grant cycle always starts with the lowest priority number and ends with the highest.
2.1.12 Miscellaneous
2.1.12.1 Receive Grants
A receive grant function is provided on device pins as a means to block packet reception for specific outputs.
This function allows the implementation of packet lossless switchover for simple switch systems. It cannot be
combined with the packet lossless switchover mechanism provided by colored packets. When a receive grant
is asserted, reception of Data Packets in the corresponding output queue is enabled. When a receive grant is
disasserted no incoming packets are stored in the corresponding output queue, regardless of their bit map.
2.1.12.2 Send Grants
Send grant pins are provided to control the outgoing flow. They indicate to the output controllers when
packets are allowed to be transmitted.
相關PDF資料
PDF描述
IBM32NPR100EXXCAB133 Microprocessor
IBM32NPR101EPXCAC133 Microprocessor
IBM37RGB524CF17A Video DAC with Color Palette (RAMDAC)
IBM37RGB524CF22A Video DAC with Color Palette (RAMDAC)
IBM39MPEGS420PBA18C
相關代理商/技術(shù)參數(shù)
參數(shù)描述
IBM32NPR100EXXCAB133 制造商:未知廠家 制造商全稱:未知廠家 功能描述:Microprocessor
IBM32NPR101EPXCAC133 制造商:未知廠家 制造商全稱:未知廠家 功能描述:Microprocessor
IBM35CPC945C03C-2 制造商:IBM 功能描述:
IBM37RGB524CF17A 制造商:未知廠家 制造商全稱:未知廠家 功能描述:Video DAC with Color Palette (RAMDAC)
IBM37RGB524CF22A 制造商:未知廠家 制造商全稱:未知廠家 功能描述:Video DAC with Color Palette (RAMDAC)