KSZ8852HLE Datasheet, PDF (15/190 Pages) Micrel Semiconductor – Two-Port 10/100 Mb/s Ethernet Switch with 8 or 16-Bit Host Interface

Micrel, Inc.
KSZ8852HLE
Acronyms
BIU
BPDU
CMOS
CRC
CUT-THROUGH SWITCH
DA
DMA
EMI
FCS
FID
IGMP
IPG
ISI
ISA
JUMBO PACKET
MAC
MDI
Bus Interface Unit
Bridge Protocol Data Unit
Complementary Metal
Oxide Semiconductor
Cyclic Redundancy Check
Destination Address
Direct Memory Access
Electromagnetic Interference
Frame Check Sequence
Frame or Filter ID
Internet Group Management
Protocol
Inter-Packet Gap
Inter-Symbol Interference
Industry Standard Architecture
Media Access Controller
Medium Dependent Interface
The host interface function that performs code conversion,
buffering, and the like required for communications to and from a
network.
A packet containing ports, addresses, etc. to make sure data being
passed through a bridged network arrives at its proper destination.
A common semiconductor manufacturing technique in which
positive and negative types of transistors are combined to form a
current gate that in turn forms an effective means of controlling
electrical current through a chip.
A common technique for detecting data transmission errors. CRC
for Ethernet is 32 bits long.
A switch typically processes received packets by reading in the full
packet (storing), then processing the packet to determine where it
needs to go, then forwarding it. A cut−through switch simply reads
in the first bit of an incoming packet and forwards the packet.
Cut−through switches do not store the packet.
The address to send packets.
A design in which memory on a chip is controlled independently of
the CPU.
A naturally occurring phenomena when the electromagnetic field of
one device disrupts, impedes or degrades the electromagnetic field
of another device by coming into proximity with it. In computer
technology, computer devices are susceptible to EMI because
electromagnetic fields are a byproduct of passing electricity through
a wire. Data lines that have not been properly shielded are
susceptible to data corruption by EMI.
See CRC.
Specifies the frame identifier. Alternately is the filter identifier.
The protocol defined by RFC 1112 for IP multicast transmissions.
A time delay between successive data packets mandated by the
network standard for protocol reasons. In Ethernet, the medium has
to be "silent" (i.e., no data transfer) for a short period of time before
a node can consider the network idle and start to transmit. IPG is
used to correct timing differences between a transmitter and
receiver. During the IPG, no data is transferred, and information in
the gap can be discarded or additions inserted without impact on
data integrity.
The disruption of transmitted code caused by adjacent pulses
affecting or interfering with each other.
A bus architecture used in the IBM PC/XT and PC/AT.
A packet larger than the standard Ethernet packet (1500 bytes).
Large packet sizes allow for more efficient use of bandwidth, lower
overhead, less processing, etc.
A functional block responsible for implementing the Media Access
Control layer which is a sub layer of the Data Link Layer.
An Ethernet port connection that allows network hubs or switches
to connect to other hubs or switches without a null−modem, or
crossover, cable. MDI provides the standard interface to a
particular media (copper or fiber) and is therefore “media
dependent”.
August 31, 2015
15
Revision 1.1