CN105511572A - Method for extending compute nodes of blade server - Google Patents
Method for extending compute nodes of blade server Download PDFInfo
- Publication number
- CN105511572A CN105511572A CN201510910610.XA CN201510910610A CN105511572A CN 105511572 A CN105511572 A CN 105511572A CN 201510910610 A CN201510910610 A CN 201510910610A CN 105511572 A CN105511572 A CN 105511572A
- Authority
- CN
- China
- Prior art keywords
- design
- node
- interlayer
- backboard
- computing
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
- G06F1/16—Constructional details or arrangements
- G06F1/18—Packaging or power distribution
- G06F1/183—Internal mounting support structures, e.g. for printed circuit boards, internal connecting means
Abstract
The invention provides a method for extending compute nodes of a blade server. According to the method, an interlayer back plate is used, reasonable structural design comprises Low-profile CPU (central processing unit) radiator design and mismatched design of memory boards on the upper layer and the lower layer, and an original single node is extended to two compute nodes in a space; two same design compute nodes are combined together through the interlayer back plate and pass by a back plate connector after signal processing. With the adoption of the interlayer back plate and the reasonable structural design, the single compute node is extended to two integrated compute nodes in the same space, and the compute performance of the nodes under the same condition can be almost doubled.
Description
Technical field
The present invention relates to a kind of interlayer backboard, be specifically related to the method for a kind of blade server expansion computing node.
Background technology
Large data, just together with cloud computing, promote the development of whole IT industry, and in the face of the demand of making rapid progress, top IT vendor all proposes oneself a set of comprehensive strategy and theory, and implements for a long time and execution.When the cloud epoch from we more and more close to time, infrastructure, application platform and application software three layers of definition have clearly had revolutionary development.Such as in infrastructure, start the different matched combined from traditional server, storage, network, started to integration and integrated direction evolution, with the cabinet of a highdensity fusion architecture, select different combinations according to the demand of different client, meet the requirement of client.
Summary of the invention
Technical assignment of the present invention is for the deficiencies in the prior art, provides a kind of blade server to expand the method for computing node.This method mainly utilizes interlayer backboard and rational structural design, realizes, in same space, single computing node is extended to 2 fusion calculation nodes, can at identical conditions node calculated performance improve nearly one times.
The technical solution adopted for the present invention to solve the technical problems is:
A kind of method of blade server expansion computing node, utilize interlayer backboard, rational structural design comprises the cpu heat design of Low-profile, and bilevel memory board dislocation design, realizes the method expanding 2 computing nodes in the space of original single node; Utilize interlayer backboard to be combined by 2 same design computing nodes, and through back panel connector after signal transacting, energy and original single computing node perfection are merged.
A computing node is by CPU, and memory bar, HDD, connection interlayer backboard connects at a high speed, the compositions such as PCH.
The I/O signal of coming from 2 computing boards is Switch by sandwich plate, then by back panel connector on backboard, like this can with the fusion mutually of single node before, do not need to change back plate design.
The step expanding 2 computing nodes in the space of a single blade node is as follows:
1) CPU radiating fin, selecting aspect ratio shorter and the design of memory bar;
2), now interlayer backboard is arranged on one of them computing node, above then another one node being buckled in;
3), interlayer backboard has individual Switch chip to be done by the signal of 2 computing nodes but to change, realize the hardware design that 2 computing nodes merge mutually with an original computing node.
The method of a kind of blade server expansion computing node of the present invention compared with prior art, the beneficial effect produced is, the present invention utilizes interlayer backboard and rational structural design, realize in same space, single computing node is extended to 2 fusion calculation nodes, can at identical conditions node calculated performance improve nearly one times.
Accompanying drawing explanation
Fig. 1 is the rough schematic diagram of the expansion computing node merging blade server;
Fig. 2 is the connection block diagram of interlayer backboard;
Fig. 3 is computing node, the schematic diagram of interlayer backboard and system dorsulum.
Embodiment
Be described in detail below below in conjunction with the method for accompanying drawing to a kind of blade server expansion computing node of the present invention.
A kind of method of blade server expansion computing node, utilize interlayer backboard, rational structural design comprises the cpu heat design of Low-profile, and bilevel memory board dislocation design, realizes the method expanding 2 computing nodes in the space of original single node; Utilize interlayer backboard to be combined by 2 same design computing nodes, and through back panel connector after signal transacting, energy and original single computing node perfection are merged.
A computing node is by CPU, and memory bar, HDD, connection interlayer backboard connects at a high speed, the compositions such as PCH.
The I/O signal of coming from 2 computing boards is Switch by sandwich plate, then by back panel connector on backboard, like this can with the fusion mutually of single node before, do not need to change back plate design.
Expand 2 computing nodes in the space of a single blade node, step is as follows:
1, the CPU radiating fin selecting aspect ratio shorter and the design of memory bar, if memory bar is short not, also it doesn't matter, because when 2 boards are buckled together time, memory bar staggers.So can be implemented in the structural design that a single node spatially expands 2 computing nodes.
2, now interlayer backboard is arranged on one of them computing node, above then another one node being buckled in.Structure as shown in Figure 3.
3, interlayer backboard has individual Switch chip to be done by the signal of 2 computing nodes but to change, and these 2 computing nodes are concerning backboard or a computing node like this.So just realize the hardware design that 2 computing nodes merge mutually with an original computing node.
Such expansion does not need any design changing cabinet and backboard, and dirigibility is high.
Claims (2)
1. the method for a blade server expansion computing node, it is characterized in that utilizing interlayer backboard, rational structural design comprises the cpu heat design of Low-profile, and bilevel memory board dislocation design, realizes the method expanding 2 computing nodes in the space of original single node; Interlayer backboard is utilized to be combined by 2 same design computing nodes, and through back panel connector after signal transacting.
2. the method for a kind of blade server expansion computing node according to claim 1, in the space that it is characterized in that a single blade node, the step of expansion 2 computing nodes is as follows:
1) CPU radiating fin, selecting aspect ratio shorter and the design of memory bar;
2), now interlayer backboard is arranged on one of them computing node, above then another one node being buckled in;
3), interlayer backboard has individual Switch chip to be done by the signal of 2 computing nodes but to change, realize the hardware design that 2 computing nodes merge mutually with an original computing node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510910610.XA CN105511572A (en) | 2015-12-10 | 2015-12-10 | Method for extending compute nodes of blade server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510910610.XA CN105511572A (en) | 2015-12-10 | 2015-12-10 | Method for extending compute nodes of blade server |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105511572A true CN105511572A (en) | 2016-04-20 |
Family
ID=55719625
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510910610.XA Pending CN105511572A (en) | 2015-12-10 | 2015-12-10 | Method for extending compute nodes of blade server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105511572A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1509041A1 (en) * | 2003-08-19 | 2005-02-23 | Medion AG | Multifunctional device for processing audio/video signals |
| CN101923380A (en) * | 2010-08-02 | 2010-12-22 | 浪潮电子信息产业股份有限公司 | Power supply design method of high-density system |
| CN102799225A (en) * | 2012-07-20 | 2012-11-28 | 浪潮电子信息产业股份有限公司 | High density server design method |
| CN203422706U (en) * | 2013-08-08 | 2014-02-05 | 北京天地超云科技有限公司 | Duel-node high-temperature energy-saving integrated server |
| CN103605404A (en) * | 2013-11-22 | 2014-02-26 | 曙光信息产业(北京)有限公司 | System with external expansion GPU (graphics processing unit) cards |
-
2015
- 2015-12-10 CN CN201510910610.XA patent/CN105511572A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1509041A1 (en) * | 2003-08-19 | 2005-02-23 | Medion AG | Multifunctional device for processing audio/video signals |
| CN101923380A (en) * | 2010-08-02 | 2010-12-22 | 浪潮电子信息产业股份有限公司 | Power supply design method of high-density system |
| CN102799225A (en) * | 2012-07-20 | 2012-11-28 | 浪潮电子信息产业股份有限公司 | High density server design method |
| CN203422706U (en) * | 2013-08-08 | 2014-02-05 | 北京天地超云科技有限公司 | Duel-node high-temperature energy-saving integrated server |
| CN103605404A (en) * | 2013-11-22 | 2014-02-26 | 曙光信息产业(北京)有限公司 | System with external expansion GPU (graphics processing unit) cards |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6101821B2 (en) | Die stacking device with split multi-hop network | |
| US9558143B2 (en) | Interconnect systems and methods using hybrid memory cube links to send packetized data over different endpoints of a data handling device | |
| US8769458B2 (en) | Prototype verification system and verification method for high-end fault-tolerant computer | |
| CN206696771U (en) | One kind is based on the road server master board of Purley platforms four | |
| US9812186B2 (en) | Reducing latency in an expanded memory system | |
| CN105187464A (en) | Data synchronization method, device and system in distributed storage system | |
| CN104504147A (en) | Resource coordination method, device and system for database cluster | |
| US20170147492A1 (en) | Method and System for Implementing Directory Structure of Host System | |
| CN109522649A (en) | Through silicon via TSV array temperature optimization method based on orthogonal test | |
| CN104199521A (en) | Blade node and extension method thereof | |
| CN104484021A (en) | Server system with expandable memory | |
| CN105893322A (en) | CPU interconnection system and realization method | |
| CN105095148A (en) | Mixed type three-dimensional on-chip network | |
| CN203689288U (en) | Complete equipment cabinet and node server of complete equipment cabinet | |
| CN104238688A (en) | Blade node and extension method thereof | |
| CN112989758B (en) | Method for synchronously resetting multiple prototype verification boards, verification system and storage medium | |
| CN102142050A (en) | Single node prototype verification system and method of high-end fault-tolerant computer | |
| CN105511572A (en) | Method for extending compute nodes of blade server | |
| Habata et al. | Hardware system of the Earth Simulator | |
| CN104065664B (en) | A kind of Cloud Server verifies system | |
| Le et al. | CSIP: A Compact Scrypt IP design with single PBKDF2 core for Blockchain mining | |
| CN206235979U (en) | A kind of overall board applied in purley Platform Servers | |
| CN203825623U (en) | Multi-core processor system based on FPGA (Field Programmable Gate Array) | |
| CN106163110B (en) | A kind of pcb board | |
| CN204680003U (en) | A kind of test board based on fusion architecture server computing unit |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160420 |
|
| RJ01 | Rejection of invention patent application after publication |