CN102694798A - Method for high-speed serial interconnection based on Hash mapping table - Google Patents
Method for high-speed serial interconnection based on Hash mapping table Download PDFInfo
- Publication number
- CN102694798A CN102694798A CN2012101492339A CN201210149233A CN102694798A CN 102694798 A CN102694798 A CN 102694798A CN 2012101492339 A CN2012101492339 A CN 2012101492339A CN 201210149233 A CN201210149233 A CN 201210149233A CN 102694798 A CN102694798 A CN 102694798A
- Authority
- CN
- China
- Prior art keywords
- hash
- data
- address
- memory bank
- list item
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Abstract
The invention provides a method for high-speed serial interconnection based on a Hash mapping table. The method comprises the following steps: describing logic by a hardware circuit by a Hash search matching principle, writing the data to be searched into a memory bank, forming a content addressable memory by combining the data and the memory bank; determining data width and data depth of the memory bank according to a port address format and the quantity of internet ports, controlling and reading the data in the HASH list item by the system through a controller after finding out corresponding address through the Hash-Key; comparing the data through a comparator after reading the data in the list item, if the data are matched, outputting the matched address and building a data transmission link.
Description
Technical field
The present invention relates to computer server or communication technical field, specifically a kind of based on the interconnected method of the high speed serialization of Hash mapping table.
Background technology
Along with the fast development of high-end server interconnection technique, on the one hand, the data bandwidth that in the internet, transmits is increasing.On the other hand, the continuous change and progress of various application in the internet, present high-end server computing node is very many; They connect through special high speed internet network; Need require the high speed internet network to provide extra treatment mechanism to realize distinguishing various service and function, like server admin node, safe access control node for the user provides more COSs and better service quality; Virtual private nodes etc., interconnected topology is very important key technology.For the increasing development trend of the data traffic that adapts to network; Guarantee the unobstructed property of internet link; High speed internet network performance has been proposed new challenge---how the technological development key network technology of fast mapping route network all relates to this problem; And based on the common inefficiency of the Hash of software mapping coupling, the method that traditional software is realized for the development of express network show unable to do what one wishes.Based on the interconnected chip of high speed serialization of Hash mapping table search matching properties fast, make it in high-end server, can be widely used.
Summary of the invention
The purpose of this invention is to provide a kind of based on the interconnected method of the high speed serialization of Hash mapping table.
The objective of the invention is to realize by following mode, search matching principle through Hash, its logic is described with hardware circuit, data to be looked into are write memory bank, both combine the constitution content addressable memory bank, comprise following functional module:
(1) high-speed serial bus;
(2) Hash offset table;
(3) Hash offset table specific store circuit;
(4) configuration register;
(5) comparator;
(6) controller.
Confirm the data width and the data depth of this memory bank according to the quantity of port address form and interconnect port; Find corresponding to behind the address through Hash_Key; System reads data in the HASH list item through controller control, compares through comparator after reading out the data in the list item, if coupling; Then export match address, set up number and pass link.
Pass through Hash table; Search the port numbers address fast; Or using data offset to represent, the Hash offset table through storage is realized searching interconnect architecture (Interconnection Fabric) based on Hash with coupling together through doubly linked list between the specific address data entries.
Hash table through storage find rapidly with the Hash table in list item be complementary, try to achieve the position of address, or more oppositely obtain interconnect architecture (Interconnection Fabric) the port numbers address of desiring to search through the address fast.
The invention has the beneficial effects as follows: the high efficiency of searching algorithm based on Hash finds out match address fast, and hardware realizes having practiced thrift interconnected time delay.
Description of drawings
Fig. 1 is a high speed serialization interconnected systems structure chart.
Embodiment
Explanation at length below with reference to Figure of description method of the present invention being done.
Of the present inventionly search matching principle through Hash, its logic is described with hardware circuit, data to be looked into write memory bank, and two textural associations become content addressable storage body (Content Addressable Memory).Design is divided into following functional module:
(1) high-speed serial bus;
(2) Hash offset table;
(3) Hash offset table specific store circuit;
(4) configuration register;
(5) comparator;
(6) controller.
Confirm to find the data width and the data depth of this memory bank corresponding to behind the address through Hash_Key according to the quantity of port address form and interconnect port, system reads data in the HASH list item through controller circuitry (6) control.Compare through comparator (5) after reading out the data in the list item, if coupling is then exported match address and accomplished the link establishment of number biography.
Embodiment 1
Pass through Hash table; Search the port numbers address fast; Or using data offset to represent, the Hash offset table through storage is realized searching interconnect architecture (Interconnection Fabric) based on Hash with coupling together through doubly linked list between the specific address data entries;
Embodiment 2
Hash table through storage find rapidly with the Hash table in list item be complementary, try to achieve the position of address, or more oppositely obtain interconnect architecture (Interconnection Fabric) the port numbers address of desiring to search through the address fast.
Except that the described technical characterictic of specification, be the known technology of those skilled in the art.
Claims (3)
1. one kind based on the interconnected method of the high speed serialization of Hash mapping table; It is characterized in that searching matching principle, its logic is described with hardware circuit, data to be looked into are write memory bank through Hash; Both combine the constitution content addressable memory bank, comprise following functional module:
(1) high-speed serial bus; (2) Hash offset table; (3) Hash offset table specific store circuit; (4) configuration register; (5) comparator; (6) controller;
Confirm the data width and the data depth of this memory bank according to the quantity of port address form and interconnect port; Find corresponding to behind the address through Hash_Key; System reads data in the HASH list item through controller control, compares through comparator after reading out the data in the list item, if coupling; Then export match address, set up number and pass link.
2. method according to claim 1; It is characterized in that through Hash table, search the port numbers address fast, or use data offset to represent; Hash offset table through storage is realized searching interconnect architecture based on Hash with coupling together through doubly linked list between the specific address data entries.
3. method according to claim 1, it is characterized in that through storage Hash table find rapidly with the Hash table in list item be complementary, try to achieve the position of address, or more oppositely obtain the interconnect architecture port numbers address of desiring to search through the address fast.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2012101492339A CN102694798A (en) | 2012-05-15 | 2012-05-15 | Method for high-speed serial interconnection based on Hash mapping table |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2012101492339A CN102694798A (en) | 2012-05-15 | 2012-05-15 | Method for high-speed serial interconnection based on Hash mapping table |
Publications (1)
Publication Number | Publication Date |
---|---|
CN102694798A true CN102694798A (en) | 2012-09-26 |
Family
ID=46860085
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2012101492339A Pending CN102694798A (en) | 2012-05-15 | 2012-05-15 | Method for high-speed serial interconnection based on Hash mapping table |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102694798A (en) |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1571352A (en) * | 2003-07-12 | 2005-01-26 | 华为技术有限公司 | Method for processing five-membered stream group in network equipment |
US20050190694A1 (en) * | 2000-04-03 | 2005-09-01 | P-Cube | Method and apparatus for wire-speed application layer classification of upstream and downstream data packets |
CN102364463A (en) * | 2011-09-19 | 2012-02-29 | 浪潮电子信息产业股份有限公司 | Hash-based method for searching CAM (central address memory) |
-
2012
- 2012-05-15 CN CN2012101492339A patent/CN102694798A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050190694A1 (en) * | 2000-04-03 | 2005-09-01 | P-Cube | Method and apparatus for wire-speed application layer classification of upstream and downstream data packets |
CN1571352A (en) * | 2003-07-12 | 2005-01-26 | 华为技术有限公司 | Method for processing five-membered stream group in network equipment |
CN102364463A (en) * | 2011-09-19 | 2012-02-29 | 浪潮电子信息产业股份有限公司 | Hash-based method for searching CAM (central address memory) |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6522663B2 (en) | Interconnection system and method using hybrid memory cube link | |
CN103444138B (en) | communication control system, switching node and communication control method | |
US20170212858A1 (en) | System for switching between a single node pcie mode and a multi-node pcie mode | |
US10162780B2 (en) | PCI express switch and computer system using the same | |
CN104871145A (en) | Memory sharing in network device | |
CN102866980B (en) | Network communication cell used for multi-core microprocessor on-chip interconnected network | |
US10547547B1 (en) | Uniform route distribution for a forwarding table | |
CN202602678U (en) | Expansion circuit with multiple CAN (Controller Area Network) channel interfaces | |
CN104079478A (en) | Method and device of packet forwarding | |
US20220398207A1 (en) | Multi-plane, multi-protocol memory switch fabric with configurable transport | |
Hu et al. | Design of all programable innovation platform for software defined networking | |
CN101222434B (en) | Storage policy control list, policy searching method and tri-state addressing memory | |
CN103023782B (en) | A kind of method and device of accessing three-state content addressing memory | |
CN105553695A (en) | IP data flow management method based on two-level bidirectional Hash list | |
CN102694798A (en) | Method for high-speed serial interconnection based on Hash mapping table | |
Mahanta et al. | Networks on chip: The new trend of on-chip interconnection | |
US20170237691A1 (en) | Apparatus and method for supporting multiple virtual switch instances on a network switch | |
Zang et al. | PROP: Using PCIe-based RDMA to accelerate rack-scale communications in data centers | |
JP5244362B2 (en) | High speed network system and related equipment | |
Kim et al. | Memory Network: Enabling Technology for Scalable Near-Data Computing | |
TW201405315A (en) | Data routing system supporting dual master apparatuses | |
CN201837923U (en) | Main board with function of switch | |
CN214959968U (en) | Data transmission system and switch with built-in data transmission system | |
EP4209916A1 (en) | Apparatus and method for cache-coherence | |
US20230055703A1 (en) | Packet processing with per flow hash key selection |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20120926 |