Information technology - High-performance parallel interface - Part 10: 6 400 Mbit/s Physical Layer (HIPPI-6400-PH)

Specifies a physical-level, point-to-point, full-duplex, link interface for reliable, flow-controlled transmission of user data at 6 400 Mbit/s per direction, across distances of up to 1 km. Characteristics of a HIPPI-6400-PH physical-layer interface include: - user data transfer bandwidth of 6 400 Mbit/s (800 MByte/s); - a full-duplex link capable of independent full-bandwidth transfers in both directions simultaneously; - four virtual circuits providing a limited multiplexing capability; - a fixed-size transfer unit, i.e., a 32-byte micropacket, for hardware efficiency; - a small transfer unit resulting in low latency for short Messages, and a component for large transfers; - credit-based flow control that prevents buffer overflow; - end-to-end, as well as link-to-link, checksums; - automatic retransmission to correct flawed data providing guaranteed, in-order, reliable, data delivery; - an a.c. coupled parallel electrical interface for driving parallel copper cable over limited distances; - a parallel electrical interface for driving a local optical interface for longer distances.

General Information

Status
Published
Publication Date
19-Mar-2001
Current Stage
PPUB - Publication issued
Start Date
20-Mar-2001
Completion Date
31-Mar-2001

Overview

ISO/IEC 11518-10:2001 - Information technology - High-performance parallel interface - Part 10: 6 400 Mbit/s Physical Layer (HIPPI-6400-PH) defines the physical-layer characteristics for a point-to-point, full‑duplex HIPPI link operating at 6 400 Mbit/s (800 MByte/s) per direction. The standard specifies a reliable, flow-controlled link for short- and long-distance connections (up to ~1 km), covering both copper and optical local physical interfaces and the electrical signalling needed to support high-throughput, low-latency data transfer.

Key topics and technical requirements

  • Data rate and topology: Fixed physical layer for 6 400 Mbit/s full‑duplex, point-to-point links.
  • Micropacket framing: Uses a fixed-size 32‑byte micropacket transfer unit for hardware efficiency and low latency on short messages.
  • Virtual circuits: Supports four virtual circuits to provide limited multiplexing across the link.
  • Flow control: Credit-based flow control prevents buffer overflow and enables sustained high throughput.
  • Reliability mechanisms: End‑to‑end and link‑to‑link checksums, plus automatic retransmission for guaranteed, in‑order, reliable delivery.
  • Signalling and encoding: Defines FRAME and CLOCK signals, skew compensation, and line encoding for parallel electrical interfaces including 4b/5b-like schemes for d.c. balance.
  • Physical interfaces: Specification of an a.c. coupled parallel electrical interface for copper cable over limited distances and a parallel electrical interface for driving a local optical interface for longer reaches (up to 1 km).
  • Operational procedures: Link initialization/reset sequences, training for skew compensation, activity monitoring, timeouts, and maintenance features.

Applications and who uses it

ISO/IEC 11518-10 (HIPPI-6400-PH) is aimed at systems requiring deterministic, very high bandwidth and low latency point-to-point links. Typical applications include:

  • High-performance computing clusters and interconnects
  • High-throughput storage arrays and SAN backplanes
  • Data acquisition systems and scientific instrumentation
  • OEMs and integrators designing HIPPI-compliant line cards, transceivers, and host interfaces

Engineers, system architects, hardware designers, and standards compliance teams use this standard to implement interoperable physical-layer interfaces and to ensure reliable, flow‑controlled data transport at 6 400 Mbit/s.

Related standards

  • ISO/IEC 11518 series (HIPPI family), including Parts 1–5 (mechanical/electrical/signalling, framing, LLC encapsulation, IPI mapping, memory interface) as part of the broader HIPPI specification suite.

Keywords: ISO/IEC 11518-10, HIPPI-6400-PH, high-performance parallel interface, 6 400 Mbit/s physical layer, credit-based flow control, micropacket, full-duplex, parallel electrical interface, reliable transmission.

Standard

ISO/IEC 11518-10:2001 - Information technology - High-performance parallel interface - Part 10: 6 400 Mbit/s Physical Layer (HIPPI-6400-PH)

English language
68 pages
sale 15% off
Preview
sale 15% off
Preview

Get Certified

Connect with accredited certification bodies for this standard

BSI Group

BSI (British Standards Institution) is the business standards company that helps organizations make excellence a habit.

UKAS United Kingdom Verified

NYCE

Mexican standards and certification body.

EMA Mexico Verified

Sponsored listings

Frequently Asked Questions

ISO/IEC 11518-10:2001 is a standard published by the International Electrotechnical Commission (IEC). Its full title is "Information technology - High-performance parallel interface - Part 10: 6 400 Mbit/s Physical Layer (HIPPI-6400-PH)". This standard covers: Specifies a physical-level, point-to-point, full-duplex, link interface for reliable, flow-controlled transmission of user data at 6 400 Mbit/s per direction, across distances of up to 1 km. Characteristics of a HIPPI-6400-PH physical-layer interface include: - user data transfer bandwidth of 6 400 Mbit/s (800 MByte/s); - a full-duplex link capable of independent full-bandwidth transfers in both directions simultaneously; - four virtual circuits providing a limited multiplexing capability; - a fixed-size transfer unit, i.e., a 32-byte micropacket, for hardware efficiency; - a small transfer unit resulting in low latency for short Messages, and a component for large transfers; - credit-based flow control that prevents buffer overflow; - end-to-end, as well as link-to-link, checksums; - automatic retransmission to correct flawed data providing guaranteed, in-order, reliable, data delivery; - an a.c. coupled parallel electrical interface for driving parallel copper cable over limited distances; - a parallel electrical interface for driving a local optical interface for longer distances.

Specifies a physical-level, point-to-point, full-duplex, link interface for reliable, flow-controlled transmission of user data at 6 400 Mbit/s per direction, across distances of up to 1 km. Characteristics of a HIPPI-6400-PH physical-layer interface include: - user data transfer bandwidth of 6 400 Mbit/s (800 MByte/s); - a full-duplex link capable of independent full-bandwidth transfers in both directions simultaneously; - four virtual circuits providing a limited multiplexing capability; - a fixed-size transfer unit, i.e., a 32-byte micropacket, for hardware efficiency; - a small transfer unit resulting in low latency for short Messages, and a component for large transfers; - credit-based flow control that prevents buffer overflow; - end-to-end, as well as link-to-link, checksums; - automatic retransmission to correct flawed data providing guaranteed, in-order, reliable, data delivery; - an a.c. coupled parallel electrical interface for driving parallel copper cable over limited distances; - a parallel electrical interface for driving a local optical interface for longer distances.

ISO/IEC 11518-10:2001 is classified under the following ICS (International Classification for Standards) categories: 35.200 - Interface and interconnection equipment. The ICS classification helps identify the subject area and facilitates finding related standards.

ISO/IEC 11518-10:2001 is available in PDF format for immediate download after purchase. The document can be added to your cart and obtained through the secure checkout process. Digital delivery ensures instant access to the complete standard document.

Standards Content (Sample)


INTERNATIONAL ISO/IEC
STANDARD
11518-10
First edition
2001-03
Information technology –
High-performance parallel interface
Part 10:
6 400 Mbit/s Physical Layer (HIPPI-6400-PH)
Reference number
INTERNATIONAL ISO/IEC
STANDARD
11518-10
First edition
2001-03
Information technology –
High-performance parallel interface
Part 10:
6 400 Mbit/s Physical Layer (HIPPI-6400-PH)
 ISO/IEC 2001
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any
means, electronic or mechanical, including photocopying and microfilm, without permission in writing from the publisher.
ISO/IEC Copyright Office Case postale 56 CH-1211 Genève 20 Switzerland
•
PRICE CODE
V
For price, see current catalogue

– 2 – 11518-10 © ISO/IEC:2001(E)
CONTENTS
FOREWORD . 6
INTRODUCTION .7
1 Scope . 8
2 Normative references. 8
3 Definitions and conventions . 9
3.1 Definitions . 9
3.2 Editorial conventions. 10
3.3 Acronyms and abbreviations . 11
4 System overview. 11
4.1 Overview . 11
4.2 Links. 12
4.3 Virtual Channels . 12
4.4 Micropacket . 13
4.5 Message . 14
4.6 FRAME and CLOCK signals. 15
4.7 Flow control . 15
4.8 Retransmission . 15
4.9 Check functions . 15
4.10 Local electrical interface (optional) . 15
4.11 Copper cable physical layer (optional) . 16
5 Service interface. 16
5.1 Overview . 16
5.2 Service primitives . 17
5.3 Sequences of primitives. 17
5.4 Data transfer service primitives. 17
5.5 Admin service primitives . 20
5.6 Control service primitives. 22
5.7 Status service primitives . 23
6 Micropacket contents . 24
6.1 Bit and byte assignments . 24
6.2 Virtual Channel (VC) selector. 26
6.3 Micropacket TYPEs. 26
6.4 Sequence number parameters . 27
6.5 Credit update parameters . 28
6.6 Check functions . 28
7 Message structure . 31
7.1 Overview . 31
7.2 MAC header. 31
7.3 LLC/SNAP header. 32
7.4 Payload . 32

11518-10 © ISO/IEC:2001(E) – 3 –
8 Source specific operations . 32
8.1 Credit update indications on Source side . 32
8.2 ACK indications on Source side . 32
8.3 ACKs and credit updates to remote end . 33
8.4 Micropacket retransmission . 33
9 Destination specific operations . 34
9.1 Link level processing . 34
9.2 Check for Message protocol errors . 34
9.3 Generating ACKs . 36
10 Signal line encoding. 36
10.1 Signal line bit assignments . 36
10.2 CLOCK and CLOCK_2 signals . 36
10.3 FRAME signal. 39
10.4 Source-side encoding for d.c. balance . 39
10.5 Destination-side decoding. 41
11 Skew compensation . 41
11.1 Training sequences. 41
11.2 Training sequence errors . 42
12 Link Reset and Initialization . 42
12.1 Overview . 42
12.2 Link Reset . 43
12.3 Initialize . 43
12.4 Hold-off timer. 45
13 Link activity monitoring and shutdown . 45
13.1 Activity monitoring . 45
13.2 Link shutdown. 45
14 Maintenance and control features . 46
14.1 Timeouts. 46
14.2 Logged events . 46
15 Local electrical interface (optional). 47
15.1 Overview . 47
15.2 Local electrical interface – Output. 49
15.3 Local electrical interface – Input . 49
15.4 Light present signal . 49
16 Copper cable interface (optional) . 51
16.1 Overview . 51
16.2 Copper cable interface – Output . 51
16.3 Copper cable interface – Input . 52
16.4 CLOCK_2 . 53
16.5 Copper cable connectors . 54
16.6 Copper cable specifications . 55
Annex A (informative) Implementation comments . 61

– 4 – 11518-10 © ISO/IEC:2001(E)
Figure 1 – System overview . 12
Figure 2 – HIPPI-6400-PH link showing signal lines . 13
Figure 3 – Logical micropacket format and naming conventions . 14
Figure 4 – Message format. 14
Figure 5 – Reverse direction control information. 16
Figure 6 – HIPPI-6400-PH service interface . 17
Figure 7 – Data transfer service primitives . 18
Figure 8 – Admin service primitives . 20
Figure 9 – Control service primitives . 22
Figure 10 – Status service primitives. 23
Figure 11 – Control bits summary. 25
Figure 12 – LCRC implementation example. 30
Figure 13 – ECRC implementation example . 30
Figure 14 – Header micropacket contents . 31
Figure 15 – Detailed ULA layout. 32
Figure 16 – 16-bit system micropacket . 40
Figure 17 – 8-bit system micropacket . 41
Figure 18 – 16-bit system training sequence . 41
Figure 19 – 8-bit system training sequence . 42
Figure 20 – Initialize and Link Reset sequences . 44
Figure 21 – Local electrical interface block diagram . 48
Figure 22 – One signal (of 12 in each direction) of the local electrical interface. 49
Figure 23 – One signal (of 23 in each direction) of the copper cable interface . 52
Figure 24 – Destination Receiver equivalent circuit . 52
Figure 25 – Receiver eye mask (differential) . 54
Figure 26 – Connecting the overall shield. 55
Figure 27 – Receptacle pin assignments . 57
Figure 28 – Receptacle . 59
Figure 29 – Cable connector . 60
Figure A.1 – Encode / decode circuit example . 61
Figure A.2 – Parallel LCRC generator example . 62
Figure A.3 – Parallel LCRC checker example . 63
Figure A.4 – Parallel ECRC example. 64
Table 1 – CRC coverages in a 128-byte Message . 16
Table 2 – Micropacket contents summary. 28
Table 3 – Signal line bit assignments in a 16-bit system . 37
Table 4 – Signal line bit assignments in an 8-bit system . 38
Table 5 – 4b/5b line coding . 40
Table 6 – Summary of timeouts . 46
Table 7 – Summary of logged events. 47
Table 8 – Local electrical signal timing at Source driver output. 50

11518-10 © ISO/IEC:2001(E) – 5 –
Table 9 – Local electrical interface, Source driver output. 50
Table 10 – Local electrical interface, Destination receiver input. 51
Table 11 – Copper cable interface. 51
Table 12 – Copper cable interface signal timing at Source driver output. 53
Table 13 – Copper cable interface, Source driver output . 53
Table 14 – Copper cable interface, Destination receiver input . 54
Table 15 – Copper cable assembly electrical specifications. 56
Table 16 – Cable layout . 58
Table A.1 – Parallel LCRC input bits . 62
Table A.2 – Parallel ECRC input bits . 63
Table A.3 – 16-bit LCRC generator equations . 65
Table A.4 – 64-bit LCRC generator equations . 66
Table A.5 – 80-bit LCRC checker equations . 67
Table A.6 – 64-bit ECRC generator / checker equations . 68

– 6 – 11518-10 © ISO/IEC:2001(E)
INFORMATION TECHNOLOGY –
HIGH-PERFORMANCE PARALLEL INTERFACE –
Part 10: 6 400 Mbit/s Physical Layer (HIPPI-6400-PH)
FOREWORD
1) ISO (International Organization for Standardization) and IEC (International Electrotechnical Commission) form
the specialized system for worldwide standardization. National bodies that are members of ISO or IEC
participate in the development of International Standards through technical committees established by the
respective organization to deal with particular fields of technical activity. ISO and IEC technical committees
collaborate in fields of mutual interest. Other international organizations, governmental and non-governmental,
in liaison with ISO and IEC, also take part in the work.
2) In the field of information technology, ISO and IEC have established a joint technical committee, ISO/IEC JTC1.
Draft International Standards adopted by the joint technical committee are circulated to national bodies for
voting. Publication as an International Standard requires approval by at least 75 % of the national bodies
casting a vote.
3) Attention is drawn to the possibility that some of the elements of this International Standard may be the subject
of patent rights. ISO and IEC shall not be held responsible for identifying any or all such patent rights.
International Standard ISO/IEC 11518-10 was prepared by subcommittee 25: Interconnection
of information technology equipment, of ISO/IEC joint technical committee 1: Information
technology.
This publication has been drafted in accordance with the ISO/IEC Directives, Part 3.
ISO/IEC 11518 consists of the following parts, under the general title Information technology –
High-Performance Parallel Interface:
– Part 1: Mechanical, electrical, and signalling protocol specification (HIPPI-PH)
– Part 2: Framing Protocol (HIPPI-FP)
– Part 3: Encapsulation of ISO/IEC 8802-2 (IEEE Std 802.2) Logical Link Control Protocol
Data Units (HIPPI-LE)
1)
– Part 4: Mapping of HIPPI to IPI device generic command sets (HIPPI-IPI)
1)
– Part 5: Memory Interface (HIPPI-MI)
– Part 6: Physical Switch Control (HIPPI-SC)
– Part 8: Mapping to Asynchronous Transfer Mode (HIPPI-ATM)
– Part 9: Serial Specification (HIPPI-Serial)
– Part 10: 6 400 Mbit/s Physical Layer (HIPPI-6400-PH)
1)
– Part 11: 6 400 Mbit/s Physical Switch Control (HIPPI-6400-SC)
1)
– Part 12: 6 400 Mbit/s Optical Specification (HIPPI-6400-OPT)
Annex A is for information only.
___________
1)
Under consideration.
11518-10 © ISO/IEC:2001(E) – 7 –
INTRODUCTION
Characteristics of a HIPPI-6400-PH physical-layer interface include:
– user data transfer bandwidth of 6 400 Mbit/s (800 MByte/s);
– a full-duplex link capable of independent full-bandwidth transfers in both directions
simultaneously;
– four virtual circuits providing a limited multiplexing capability;
– a fixed-size transfer unit, i.e., a 32-byte micropacket, for hardware efficiency;
– a small transfer unit resulting in low latency for short Messages, and a component for
large transfers;
– credit-based flow control that prevents buffer overflow;
– end-to-end, as well as link-to-link, checksums;
– automatic retransmission to correct flawed data providing guaranteed, in-order, reliable,
data delivery;
– an a.c. coupled parallel electrical interface for driving parallel copper cable over limited
distances;
– a parallel electrical interface for driving a local optical interface for longer distances.

– 8 – 11518-10 © ISO/IEC:2001(E)
INFORMATION TECHNOLOGY –
HIGH-PERFORMANCE PARALLEL INTERFACE –
Part 10: 6 400 Mbit/s Physical Layer (HIPPI-6400-PH)
1 Scope
This part of ISO/IEC 11518 specifies a physical-level, point-to-point, full-duplex, link interface
for reliable, flow-controlled transmission of user data at 6 400 Mbit/s per direction, across
distances of up to 1 km. A parallel copper cable interface for distances of up to 40 m is
specified. Connections to a separate longer-distance optical interface are provided. Small
fixed-size micropackets provide an efficient, low-latency structure for small transfers, and a
component for large transfers.
Specifications are included for:
– automatic retransmission to correct flawed data;
– the format of a small data transfer unit called a micropacket;
– a message structure that includes routing information for network applications;
– end-to-end, as well as link-to-link, checksums;
– the timing requirements of the parallel signals;
– a parallel interface using copper coaxial cable;
– connections to a separate local optical interface;
– a link-level protocol tuned for a maximum distance of 1 km.
2 Normative references
The following normative documents contain provisions which, through reference in this text,
constitute provisions of this part of ISO/IEC 11518. For dated references, subsequent
amendments to, or revisions of, any of these publications do not apply. However, parties to
agreements based on this part of ISO/IEC 11518 are encouraged to investigate the possibility
of applying the most recent edition of the normative documents indicated below. For undated
references, the latest edition of the normative document referred to applies. Members of IEC
and ISO maintain registers of currently valid International Standards.
ISO/IEC TR 8802 (all parts), Information technology – Telecommunications and information
exchange between systems – Local and metropolitan area networks – Specific requirements
ISO/IEC TR 8802-1:1997, Information technology – Telecommunications and information
exchange between systems – Local and metropolitan area networks – Specific requirements –
Part 1: Overview of Local Area Network Standards
ISO/IEC 8802-2:1998, Information technology – Telecommunications and information
exchange between systems – Local and metropolitan area networks – Specific requirements –
Part 2: Logical link control
ISO/IEC 15802-3:1998, Information technology – Telecommunications and information
exchange between systems – Local and metropolitan area networks – Common specifications
– Part 3: Media Access Control (MAC) Bridges

11518-10 © ISO/IEC:2001(E) – 9 –
3 Definitions and conventions
3.1 Definitions
For the purposes of this part of ISO/IEC 11518, the following definitions apply.
3.1.1
acknowledge (ACK)
confirmation that the Destination has received the micropacket without errors
3.1.2
administrator
station management entity providing external management control
3.1.3
credit
credit corresponds to one micropacket's worth of buffer space available in the Destination's
VC buffer
3.1.4
destination
receiving end of a physical link
3.1.5
element
component of a HIPPI-6400 system that is able to receive, process, and send Admin
micropackets
3.1.6
final destination
end device that receives, and operates on, the data payload portion of the micropackets
This is typically a host computer system, but may also be a non-transparent translator, bridge,
or router.
3.1.7
link
full-duplex connection between HIPPI-6400-PH devices
3.1.8
log
act of making a record of an event for later use
3.1.9
message
ordered sequence of one or more micropackets that have the same VC, Originating Source,
and Final Destination
Messages are the basic transfer unit between an Originating Source and a Final Destination.
The first micropacket of a Message is a Header micropacket. The last micropacket, which may
also be the first micropacket, has the TAIL bit set (see 4.5).
3.1.10
micropacket
basic transfer unit, between a Source and Destination, consisting of 32 data bytes and 64 bits
of control information (see 4.4)

– 10 – 11518-10 © ISO/IEC:2001(E)
3.1.11
next-layer
protocols above the service interface
These could be implemented in hardware or software, or they could be distributed between
the two.
3.1.12
optional
characteristics that are not required by HIPPI-6400-PH
However, if any optional characteristic is implemented, it shall be implemented as defined in
HIPPI-6400-PH.
3.1.13
originating source
end device that generates the data payload portion of the micropackets
This is typically a host computer system, but may also be a non-transparent translator, bridge,
or router.
3.1.14
source
sending end of a physical link
3.1.15
station management (SMT)
supervisory entity that monitors and controls the HIPPI-6400-PH entity
3.1.16
syndrome
value (should be zero if no error) obtained by exclusive ORing the calculated CRC value with
the CRC value received with the micropacket
3.1.17
Universal LAN MAC Address (ULA)
logical address stored in a Source or Destination field that uniquely identifies an Originating
Source or Final Destination
The ULA conforms to the 48-bit MAC address specified by ISO/IEC 8802-1.
3.1.18
Virtual Channel (VC)
one of four logical paths within each direction of a single link
3.2 Editorial conventions
In this part of ISO/IEC 11518, certain terms that are proper names of signals or similar terms
are printed in upper case to avoid possible confusion with other uses of the same words (e.g.,
FRAME). Any lower case uses of these words have the normal technical English meaning.
A number of conditions, sequence parameters, events, states, or similar terms are printed
with the first letter of each word in upper case and the rest lower case (e.g., Block, Source).
Any lower case uses of these words have the normal technical English meaning.
The word shall, when used in this part of ISO/IEC 11518, states a mandatory rule or
requirement. The word should, when used in this part of ISO/IEC 11518, states a
recommendation.
11518-10 © ISO/IEC:2001(E) – 11 –
3.2.1 Binary notation
Binary notation is used to represent relatively short fields. For example a two-bit field
containing the binary value of 10 is shown in binary format as b'10'.
3.2.2 Hexadecimal notation
Hexadecimal notation is used to represent some fields. For example a two-byte field
containing a binary value of b'1100010000000011' is shown in hexadecimal format as x'C403'.
3.3 Acronyms and abbreviations
ACK acknowledge indication
CR credit amount parameter
CRC cyclic redundancy check
DSAP Destination Service Access Protocol
ECRC end-to-end CRC
HIPPI High-Performance Parallel Interface
K kilo (2 or 1024)
LCRC link CRC
LLC Logical Link Control
lsb least significant bit
M mega (10 )
MAC Media Access Control
ms milliseconds
msb most significant bit
ns nanoseconds
ps picoseconds
RSEQ receive sequence number
SMT station management
SNAP SubNetwork Access Protocol
SSAP Source Service Access Protocol
TSEQ transmit sequence number
ULA Universal LAN Address
VC Virtual Channel
VCR Virtual Channel Credit selector
µs microseconds
Ω ohms
4 System overview
4.1 Overview
This clause provides an overview of the structure, concepts, and mechanisms in HIPPI-6400-PH.
Figure 1 shows an example HIPPI-6400 system.

– 12 – 11518-10 © ISO/IEC:2001(E)
4.2 Links
HIPPI-6400-PH defines a point-to-point physical link for transferring micropackets. The
physical links, as shown in figure 2 between a local end and a remote end, are bi-directional.
The logical links are simplex, i.e., the data inbound and outbound are completely separate.
Some control information, e.g., credit, flows in the reverse direction, and it is included in the
micropackets flowing in the reverse direction. This is why the physical links must be bi-
directional with information flowing in both directions simultaneously.
A link is composed of two Sources that transmit information, and two Destinations that receive
information. Each end of a link has a Source and a Destination.
The data path is 16 bits wide for the copper implementation, and is 8 bits wide for a fibre
implementation. The control path is one-fourth the width of the data path, e.g., the control
path for the copper implementation is 4 bits wide. CLOCK, CLOCK_2, and FRAME are
individual signals carried on separate conductors. The CLOCK_2 signal is only used in 16-bit
systems.
4.3 Virtual Channels
Four Virtual Channels, VC0, VC1, VC2, and VC3 are available in each direction on each link.
The VCs are assigned to specific Message sizes and transfer methods.
All of the micropackets of a Message are transmitted on a single VC, i.e., the VC number
does not change as the micropackets travel from the Originating Source to the Final
Destination over one or more links. Messages to a Final Destination are delivered in order on
a single VC. Multiple messages may be out of order if sent over different VCs-even if the VCs
are in the same physical link. The VCs provide a multiplexing mechanism that can be used to
prevent a large Message from Blocking a small Message until the large Message has
completed.
S S
D D
HIPPI- 6400
HIPPI- 6400
HIPPI- 6400
Node
Sw itch
Node
D S D
S
D S D S
= HIPPI-6400
S D S D
= Other
T ranslation T ranslation
D = Destination
Func tion Func tion
D = Final Destination
S
= Source
Othe r Media
H IPPI-800
Node
= Originating Source
Node
S
Figure 1 – System overview
11518-10 © ISO/IEC:2001(E) – 13 –
Remote end
Local end
Source Destination
16(8)
DATA
DATA
4(2)
CONTROL
CONTROL
1(1)
FRAME
FRAME
1(1)
CLOCK
CLOCK
1(0)
CLOCK_2
CLOCK_2
Destination Source
16(8)
DATA
DATA
4(2)
CONTROL
CONTROL
1(1)
FRAME
FRAME
1(1)
CLOCK
CLOCK
1(0)
CLOCK_2 CLOCK_2
(Numbers in parenthesis are for an 8-bit system.
CLOCK_2 is only used in 16-bit systems.)
Figure 2 – HIPPI-6400-PH link showing signal lines
4.4 Micropacket
Micropackets are the basic transfer unit from Source to Destination on a link. As shown in
figure 3, a micropacket is composed of 32 data bytes and 64 bits of control information. At
6 400 Mbit/s a micropacket is transmitted every 40 ns, with Null micropackets transmitted
when other micropackets are not available. Credit and retransmit operations are performed on
a micropacket basis.
The 64 bits of control information in each micropacket includes parameters for:
– selecting a VC;
– detecting missing micropackets;
– denoting the types of information in the micropacket;
– marking the last micropacket of a Message;
– signaling that the Message was truncated at its originator, or damaged en-route;
– passing credit information from the Destination to the Source;
– Link-level and end-to-end checksums.

– 14 – 11518-10 © ISO/IEC:2001(E)

32 Data bytes ( 256 bits)
Data byte DB00 Data byte DB01
d00.7 d00.5 d00.3 d00.1 d01.7 d01.5 d01.3 d01.1
d00.6 d00.4 d00.2 d00.0 d01.6 d01.4 d01.2 d01.0
Data byte DB30 Data byte DB31
7 6 5 4 3 2 1 0
2 2 2 2 2 2 2 2
d30.7 d30.5 d30.3 d30.1 d31.7 d31.5 d31.3 d31.1
d30.6 d30.4 d30.2 d30.0 d31.6 d31.4 d31.2 d31.0
64 Control bits
........ 7 6 5 4 3 2 1 0
2 2 2 2 2 2 2 2
c63 c61 c59 c57 c15 c13 c11 c09 c07 c05 c03 c01
c62 c60 c58 c56 c14 c12 c10 c08 c06 c04 c02 c00
Naming conventions:
Data bytes are labelled capital DB and a two-digit number, e.g., DB00.
In a parameter that uses multiple Data bytes, the most-significant byte is the lowest-numbered byte.
Data bits are labelled lower case d, a two-digit byte number, and a one-digit bit number, e.g., d31.7.
Within each Data byte, the most-significant bit is the highest-numbered bit, i.e., dnn.7.
Control bits are labelled lower case c and a two-digit number, e.g., c00.
In a parameter that uses multiple Control bits, the most-significant bit is the highest-numbered bit.
Figure 3 – Logical micropacket format and naming conventions
4.5 Message
As shown in figure 4, a Message is an ordered sequence of one or more micropackets that
have the same VC, Originating Source, and Final Destination. The first micropacket of a
Message, i.e., the Header micropacket, contains information used to route through a HIPPI-
6400 fabric. The last micropacket of the Message is marked with the TAIL bit.
1 Header information, Bytes 0 - 7 c63–c00
2 Bytes 8 - 39 of Message data c63–c00
Bytes 40 - 71 of Message data
3 c63–c00
. .
. .
. .
. .
. .
. .
. .
n Last bytes of Message data c63–c00
Micropacket transmission order
Figure 4 – Message format
11518-10 © ISO/IEC:2001(E) – 15 –
4.6 FRAME and CLOCK signals
The FRAME signal, carried on a separate signal line, marks a micropacket's beginning. Both
edges of either the CLOCK or CLOCK_2 signals, also carried on separate signal lines, are
used for strobing the data. The data, control, and FRAME signals from a Source are
synchronous with that Source's CLOCK and CLOCK_2 signals. The CLOCK rate is dependent
on the width of the data bus, e.g., a 16-bit data bus utilizing 4b/5b encoding requires the
CLOCK line to run at 250 MHz and each data and control line may transition every 2 ns.
4.7 Flow control
Link-level credit-based flow control is used between a Source and Destination. As shown in
figure 5, the credits are assigned on a VC basis, i.e., VC0's credits are separate from VC1's
credits. The Destination end of a link grants credits to match the number of free receive
buffers for a particular VC. The Source end of the link consumes credits as it moves
micropackets from the VC Buffers to the Output Buffer. Note that flow control is on a link
basis, i.e., hop-by-hop.
4.8 Retransmission
Retransmission is performed to correct flawed micropackets (see 8.4). Go-back-N
retransmission is used, i.e., if an error is detected, then the flawed micropacket, and all
micropackets transmitted after it, are retransmitted. The CRCs in each micropacket are
checked at the Destination side of a link; at the Input Buffer in figure 5. Correct micropackets
are acknowledged, flawed micropackets are discarded. Note that retransmission is
independent of the VC used, and also independent of the credit information, i.e.,
retransmission occurs between the Output and Input Buffers in figure 5 while VC and credit
information pertains only to the VC Buffers. Retransmission is on a link basis, i.e., hop-by-
hop.
4.9 Check functions
As shown in table 1, two 16-bit cyclic redundancy checks (CRCs) are used, and they use
different polynomials. The end-to-end CRC (ECRC) covers the data bytes of all of the
micropackets in a Message, i.e., the Header micropacket and all of the Data micropackets (if
any) up to this point in a Message. The ECRC does not cover the control bits. The ECRC is
unchanged from the Originating Source to the Final Destination. The ECRC is accumulated
over an entire Message, i.e., it is not reinitialized for intermediate Data micropackets (see
6.6.3). Note that in table 1, the second micropacket's ECRC covers the information in the first
and second micropacket; the third micropacket's ECRC covers the information in the first,
second, and third micropacket, etc.
The link CRC (LCRC) covers all of the data and control bits of a micropacket, with the
exception of itself (see 6.6.2). The LCRC is initialized for each micropacket, and must be
calculated fresh for each link since other control fields change.
The combination of two 16-bit CRCs provides a stronger check than a single 16-bit CRC for
link-level checking of individual micropackets. In addition, the 16-bit ECRC provides checking
over a whole Message.
4.10 Local electrical interface (optional)
The optional local on-board electrical interface (see clause 15) provides a connection to a
1)
separately specified optical interface (see ISO/IEC 11518-12, HIPPI-6400-OPT) for longer
distances. Note that the TSEQ and RSEQ parameter sizes in this part of ISO/IEC 11518
support full speed operation at distances up to 1 km.
___________
1)
ISO/IEC 11518-12, Information technology – High Performance Parallel Interface – 6 400 Mbit/s Optical
Interface (HIPPI-6400-OPT) (under consideration).

– 16 – 11518-10 © ISO/IEC:2001(E)
4.11 Copper cable physical layer (optional)
The optional HIPPI-6400-PH copper cable variant (see clause 16) uses a cable with 46
conductor pairs, 23 in each direction, and an overall shield. The maximum length is
dependent upon the quality of the cable. The signals are a.c. coupled to the cable to
accommodate some difference in the ground potential between the equipment.
Credits are consumed as a
Destination
Source
micropacket moves from the
VCn Buffer to the Output Buffer.
VC0 Buffer
VC0 Buffer
Input
Output
Buffer
Buffer
VC1 Buffer
VC1 Buffer
VC2 Buffer
VC2 Buffer
TSEQ
ACK(seq)
RSEQ
VC3 Buffer
VC3 Buffer
ACKs are generated independent of the VC
number, and sent to the Source in the reverse
direction micropacket control information.
credit (VC,amount)
Credits are generated, on a VC basis when data
exits from the VC buffer, and sent to the Source in
the reverse direction micropacket control information.
Figure 5 – Reverse direction control information
Table 1 – CRC coverages in a 128-byte Message
Data Bytes
Micropacket ECRC LCRC
DB00 – DB31
number coverage coverage
contents
1 Header, Bytes – 7 Header, Bytes 0 – 7 Header, Bytes 0 – 7, c00 – c47
2 Bytes 8 – 39 Header, Bytes 0 – 39 Bytes 8 – 39, c00 – c47
3 Bytes 40 – 71 Header, Bytes 0 – 71 Bytes 40 – 71, c00 – c47
4 Bytes 7 – 103 Header, Bytes 0 – 103 Bytes 72 – 103, c00 – c47
5 Bytes 10 – 135 Header, Bytes 0 – 135 Bytes 104 – 135, c00 – c47
NOTE For a 128-byte Message, bytes 128-135 would be pad bytes containing x'00'.
5 Service interface
5.1 Overview
This clause specifies the services provided by HIPPI-6400-PH. The intent is to allow next-
layers to operate correctly with this HIPPI-6400-PH. How many of the services described
herein are chosen for a given implementation is up to that implementor; however, a set of
HIPPI-6400-PH services shall be supplied sufficient to satisfy the next-layer(s) being used.
The services as defined herein do not imply any particular implementation, or any interface.

11518-10 © ISO/IEC:2001(E) – 17 –
Figure 6 shows the relationship of the HIPPI-6400-PH interfaces.
next-layer
Transfer  Admin
Station
HIPPI-6400
management
-PH
(SMT)
Figure 6 – HIPPI-6400-PH service interface
5.2 Service primitives
HIPPI-6400-PH service primitives are of four types.
– Request primitives are issued by a service user to initiate a service provided by the HIPPI-
6400-PH.
– Confirm primitives are issued by the HIPPI-6400-PH to acknowledge a Request.
– Indicate primitives are issued by the HIPPI-6400-PH to notify the service user of a local
event. This primitive is similar in nature to an unsolicited interrupt. Note that the local
event may have been caused by a service Request.
– Response primitives are issued by a service user to acknowledge an Indicate.
5.3 Sequences of primitives
The order of execution of service primitives is not arbitrary. Logical and time sequence
relationships exist for all described service primitives. Time sequence diagrams are used to
illustrate a valid sequence. Other valid sequences may exist. The sequence of events
between peer users across the user/provider interface is illustrated. In the time sequence
diagrams, the HIPPI-6400-PH users are depicted on either side of the vertical bars while the
HIPPI-6400-PH acts as the service provider.
In this part of ISO/IEC 11518, a second Request primitive of the same name for the same VC
shall not be issued until the Confirm for the first request is received. Likewise, a second
Indicate primitive of the same name for the same VC shall not be issued until the Response
for the first Indicate is received. In addition, since TRANSFER and ADMIN operations of the
same type share VC resources, only one at a time is allowed on the same VC.
5.4 Data transfer service primitives
These primitives, as shown in figure 7, shall be used to transfer next-layer data from an
Originating Source next-layer to a Final Destination next-layer. The next-layer data shall be
carried in a Message, with HIPPI-6400-PH MAC and IEEE 802.2 LLC/SNAP headers
preceding the next-layer payload data (see figures 4 and 14, and clause 7). The next-layer
data shall immediately follow the LLC/SNAP header.
While figure 7 shows the data being transferred after the 64_TRANSFER.Confirm is issued,
this ordering is not mandatory.

– 18 – 11518-10 © ISO/IEC:2001(E)
Originating HIPPI- Final
Source ULP 6400-PH Destination ULP
64_TRANSFER
.Request
64_TRANSFER
.Confirm
64_TRANSFER
.Indicate
64_TRANSFER
.Response
Figure 7 – Data transfer service primitives
5.4.1 64_TRANSFER.Request
This primitive is issued by the Originating Source's next-layer to request a data transfer.
Semantics – 64_TRANSFER.Request (
D_ULA,
S_ULA,
VCn,
LLC/SNAP,
Length,
Data)
The D_ULA (Destination address) shall be placed directly in the MAC header (see 7.2).
The S_ULA (Source address), if allowed by S_ULA_Allowed = true (see 5.6.1), shall be
placed directly in the MAC header (see 7.2). If S_ULA_Allowed = false, then the HIPPI-6400-
PH entity shall use its native S_ULA address. Note that by allowing the next-layer to specify
the Source address, a server can use a "spoof" address, e.g., to provide a broadcast service.
Whether the next-layer is allowed to set the S_ULA or not is controlled by the local station
management entity through the S_ULA_Allowed flag (see 5.6.1).
VCn shall be the Virtual Channel (see 6.2) that the message shall be sent on.
LLC/SNAP shall be the LLC/SNAP header, including the appropriate EtherType specifying the
data type (see 7.3).
Length shall specify the number of bytes of next-layer payload data. Note that the length
parameter in the MAC header (see clause 7) M_len = Length + 8 to account for the LLC/SNAP
header.
Data shall be the next-layer payload data.
Issued – The Originating Source next-layer issues this primitive to the HIPPI-6400-PH entity
to request the transfer of the next-layer payload data to the Final Destination. Note that one
64_TRANSFER.Request or one 64_ADMIN.Request may be issued for each of the four VCs
before receiving a Confirm for any of them, i.e., the .Confirm / .Request exchange is on a per-
VC basis. For example, a 64_TRANSFER.Request for VC1 shall not be issued if a
64_ADMIN.Request is in progress on VC1.
Effect – The HIPPI-6400-PH entity shall accept the data for transmission and build the
Message with the appropriate MAC and LLC/SNAP headers. If the Message size violates the
Virtual Channel limitations (see 6.2), then this transfer request shall be rejected (see 5.4.2);
otherwise, the Message shall be sent. If the Message does not end on a micropacket
boundary, then padding shall be provided (see clause 7).

11518-10 © ISO/IEC:2001(E) – 19 –
5.4.2 64_TRANSFER.Confirm
This primitive acknowledges the 64_TRANSFER.Request from the Originating Source next-
layer.
Semantics –
...

Questions, Comments and Discussion

Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.

Loading comments...